Functional Relationships Among Rubisco Family Members

FUNCTIONAL RELATIONSHIPS AMONG RUBISCO FAMILY MEMBERS

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy

in the Graduate School of The Ohio State University

Jaya Singh, Ph.D.

*****

The Ohio State University

2008

Dissertation Committee:

Dr. F. Robert Tabita, Advisor Approved by

Dr. Iris Meier

Dr. Birgit E. Alber

Dr. Patrice Hamel

______

Advisor

Graduate Program of Plant Cellular and

Molecular Biology

ABSTRACT

Ribulose 1, 5-bisphosphate carboxylase/oxygenase (RubisCO), the rate-limiting

enzyme of the Calvin-Benson-Bassham (CBB) reductive pentose phosphate pathway,

catalyzes the assimilation of atmospheric CO2 into organic matter and is responsible for

the bulk of the reduced organic matter found on earth. The beginning of this decade was

marked by the discovery of a new family of proteins, the RubisCO-like proteins (RLP),

which are structural homologs of RubisCO. From studies carried out so far, RLPs seem

to lack the ability to catalyze CO2 fixation despite sharing structural homology with bonafide RubisCOs. The RLPs from Chlorobium tepidum, Bacillus subtilis, Geobacillus kaustophilus, and Microcystis aeruginosa have been shown to be involved in sulfur

metabolism. Whereas the precise function of C. tepidum RLP is unknown, the B. subtilis,

G. kaustophilus, and M. aeruginosa RLPs function as enolases in a methionine salvage

pathway (MSP). Interestingly, the RubisCO from R. rubrum has also been shown to

weakly catalyze an enolase reaction similar to RLP of B. subtilis.

In the current study, the functional roles of diverse RLPs and their relation to

RubisCO catalysis have been examined in three organisms containing diverse RLP

lineages; i.e., C. tepidum (from the Photo clade of RLP), R. rubrum (from the DeepYkrW

clade) and Rhodopseudomonas palustris (which has RLPs from both the Photo and

DeepYkrW clades). Earlier it was shown that disruption of the RLP gene of C. tepidum

iii resulted in a pleiotropic phenotype. The current study indicates that the lack of RLP in C.

tepidum results in complex changes that appear to not be reversed by complementation

of the wild-type RLP gene on a plasmid vector. One of the phenotypes, i.e., the

overexpression of oxidative stress-response proteins Tsa (thiol specific anti-oxidant) and

SOD (superoxide dismutase), was examined further and it was shown that regulation of

these genes is not under control of the conventional regulators, Fur and PerR. Studies

with R. rubrum and R. palustris showed the presence of a functional MSP in these organisms, surprisingly, under anaerobic conditions. Further studies with R. rubrum showed the involvement of endogenous RLP and RubisCO in MSP under aerobic and anaerobic conditions, respectively. However, unlike B. subtilis RLP, the R. rubrum RLP

does not appear to catalyze the enolization of DK-MTP-1-P. It uses another intermediate

of MSP, MTRu-1-P, as a substrate and converts it into a different compound via a novel

chemistry. It had earlier been reported that the R. rubrum RubisCO-deletion strain could grow photoautotrophicaly using thiosulfate and sulfide as electron donors. In order to address this anomaly, the presence of the ethylmalonyl-CoA pathway for acetate assimilation was investigated. In the process, crotonyl-CoA carboxylase/reductase, a key enzyme in the pathway, was shown to be present.

ACKNOWLEDGMENTS

I would like to express my profound gratitude to Dr. F. Robert Tabita, my thesis advisor, for his immense support, patience and encouragement throughout the course of my graduate work.

I also owe thanks to my committee members Prof. Iris Meier, Dr. Patrice Hamel and my special thanks to Dr. Birgit Alber, for all her help and advice. I also want to thank her for providing crotonyl-CoA, CcR antibody and helping me with the enzyme assays. I would also like to thank Dr. Richard Swenson for serving on my candidacy exam committee and his advice.

I would also like to thank Dr. Tom Hanson, who introduced me to the world of RubisCO-like proteins and for teaching me all the basic techniques.

I am grateful for the support and help I got from my lab members both past and present, which was essential for me through my thesis.

I want to thank our collaborators Dr. John A. Gerlt and Heidi Imker at University of Illinois, Urbana Champaign, for in vitro characterization of the R. rubrum RLP reaction. I also want to thank Dr Thomas Pochapsky at Brandeis University for the generous gift of RLP substrate analog for the initial studies and Dr. R. S. Pareli at the Ohio State University for the synthesis of the analog.

I would also like to thank my parents for giving me strong basic education and encouragement to pursue my interests. My special thanks to my brother Rahul for providing inspiration and initial support to have the courage to move away from my home country and follow my dream. I also want to thank my mother in law Mrs. Sundara Satagopan for all her support and taking care of me and my son during the last stage of my Ph. D.

Finally, I wish to thank my husband Dr. Sriram Satagopan for providing critical advice on my work as a co-worker and moral support and encouragement at personal level. This work would not have been possible with out his support.

VITA

April 21, 1976...………………….…………..Born – Kanpur, India

1997 ………………………………………….B.Sc. Biology, A. N. D. College, Kanpur, India

2000 …………………..………...... M. Sc, G. B. Pant University, India

2001- Present………………………………..Graduate Teaching and Research Associate,

The Ohio State University

PUBLICATIONS

Tabita, F. R., T. E. Hanson, H. Li, S. Satagopan, J. Singh, and S. Chan. 2007.

Function, structure, and evolution of the RubisCO-like proteins and their RubisCO homologs. Microbiol Mol Biol Rev.71(4):576-99.

FIELDS OF STUDY

Major Field: Plant Cellular and Molecular Biology

TABLE OF CONTENTS

Abstract...... iii

Acknowledgements...... v

Vita...... vi

List of Tables...... ix

List of Figures...... ix

List of Abbreviations...... ix

Chapters:

1. Introduction...... 1

2. Changes elicited by disruption of the RubisCO-Like protein gene in the green

sulfur bacterium Chlorobium tepidum...... 16

Introduction...... 16

Materials and Methods...... 19

Results...... 28

Discussion...... 42

3. Role of RubisCO and the RubisCO-Like protein in 5-methylthioadenosine

metabolism in nonsulfur purple bacteria...... 44

Introduction...... 44

Materials and Methods...... 50

Results...... 60

vii Discussion...... 82

4. Photoautotrophic growth of a Rhodospirillum rubrum RubisCO mutant………89

Introduction...... 89

Materials and Methods...... 93

Results...... 95

Discussion...... 99

5. Recapitulation and future directions………………….…………….……………..101

Summary of work performed thus far…...... 101

Suggested Future Experiments...... ……………………………………..…….107

References...... 110

viii

LIST OF TABLES

Table Page no.

2.1 Plasmids used in the study…………………………………………………..26

2.2 Strains used in the study…………………………………………………….27

3.1 Plasmids used in the study………………………………………………….55

3.2 Strains used in the study…………………………………………………….56

3.3 Primers used for complementation studies………………………..………57

3.4 MTA dependent growth of R. palustris strains……………………………..73

3.5 MTA dependent growth of R. rubrum strains complemented with R.

Palustris cbbM…………………………………………………….…..………76

3.6 Specific activity of B. subtilis and R. rubrum RLPs as DK-H 1-P

enolase………………………………………………………………………....78

4.1 Summary of genome search…………………………………………………90

4.2 CcR activity of R. rubrum strains grown with different carbon sources..98

List of Figures

Figure Page no.

1.1 Carboxylation and oxygenation reaction of RubisCO…………....…………3

1.2 Phylogenetic tree of RubisCO family of proteins…………………..………..5

1.3 Alignment showing conservation of RubisCO active site residues in

RubisCO family members……………………………………………..………8

ix 1.4 Local gene conservation near RLP genes of different lineages…………10

1.5 Comparison of RubisCO and RLP enolase reaction mechanism………..13

2.1 SDS-PAGE gel showing Tsa/SOD hyper accumulated band in Ω::RLP

Absence in Ω::RLP/tsa- strain………………………………………………..29

2.2 SDS-PAGE gel showing protein profile or perR mutants…………………30

2.3 2D-SDS-PAGE gel showing absence of Tsa/SOD hyper accumulation in

perR mutants……………………………………………………………….….31

2.4 Figure of showing pigmentation defect in fur mutants…………………….33

2.5 Absorption spectra showing 4-5 nm blue shift in vivo absorption peak for

BChl c in fur mutants……………………………………………………….…34

2.6 SDS-PAGE gel showing protein profile or fur mutants……………………35

2.7 2D-SDS-PAGE gel showing absence of Tsa/SOD hyper accumulation in

fur mutants…………………………………………………………………….36

2.8 Map of vector used for complementation studies in C. tepidum…………38

2.9 SDS-PAGE gel and western blot showing successful transfer and

expression of RLP gene into Ω::RLP strain by conjugation………………40

2.10 SDS-PAGE gel and western blot showing successful transfer and

expression of RLP gene into Ω::RLP strain by transferring whole RLP

operon by conjugation………………………………………………………..41

3.1 Phylogenetic tree of RLP family……………………………………………..45

3.2 Local gene conservation near RLP genes of DeepYkrW lineage……….46

3.3 Methionine Salvage Pathway………………………………………………..49

3.4 Methionine and MTA dependent growth of four nonsulfur purple

bacteria………………………………………………………………………...61

x 3.5 Southern blot showing disruption of RLP gene……………………………63

3.6 MTA dependent growth curve of R. rubrum strains showing inability of

RLP mutants to grow under aerobic condition………………………..……65

3.7 MTA dependent growth curve of R. rubrum strains showing inability of

RubisCO mutants to grow under anaerobic condition…………………….67

3.8 MTA dependent growth of RubisCO-/RLP- strain complemented with

different RubisCO and RLP genes ………………..………………………..69

3.9 MTA dependent growth of RubisCO-/RLP- strain complemented with R.

rubrum RubisCO gene ……………………………………………………….71

3.10 MTA dependent growth of RubisCO-/RLP- strain complemented with R.

palustris wild type and mutants in RubisCO gene ...……………………..72

3.11 MTA dependent growth of R. palustris wild type …………………..……..75

3.12 MTA dependent growth of R. rubrum wild type and mtrpI – strain….……77

3.13 SDS-PAGE gel showing R. rubrum RLP purification…………………..…79

3.14 Figure showing enolase reaction Catalyzed by B. subtilis RLP and novel

reaction catalyzed by R. rubrum RLP………………………………………81

4.1 Ethylmalonyl-CoA pathway…………………………………………………..92

4.2 Growth of R. rubrum wild type and RubisCO mutant strain in autotrophic

PF-7 medium…………………………………………………………………..96

4.3 Western blot showing absence of RubisCO protein in RubisCO mutant

strain of R. rubrum…………………………………………………………….97

5.1 Local gene conservation near RLP genes of lineage IV-DeepYkrW

showing presence of a gene encoding for cupin2 super family protein.108

LIST OF ABBREVIATIONS

CBB Calvin-Benson-Bassham

CP Chlorobium Plating

DK-MTP-1-P 2,3- diketo- 5 - methylthiopentyl-1-phosphate

DNA Deoxyribonucleic acid

Fur Ferric uptake regulator

HK-MTP 1-P 2-hydroxy-3-keto-5-methylthiopentenyl-1-phosphate

KMTB 2-keto - 4- methylthiobutyrate

LB Luria-Bertani

MCS Multiple Cloning Site

MSP Methionine Salvage Pathway

MTA Methylthioadenosine

OD Optical Density

OM Ormerod’s Medium

ORF Open Reading Frame

PA Photoautotrophic

PH Photoheterotrophic

PCR Polymerase Chain Reaction

xii perR Peroxide Regulator

PYE Peptone Yeast Extract

RLP RubisCO Like Protein

RNA Ribonucleic acid

RPM Revolution Per Minute

RTCA Reductive Tricarboxylic Acid

RT-PCR Reverse Transcription-Polymerase Chain Reaction

RuBP Ribulose bis phosphate

RubisCO Ribulose bis phosphate carboxylase/oxygenase

1D-SDS-PAGE 1-Dimensional-Sodium Dodecyl Sulfate Polyacrylamide

Gel Electrophoresis

2D-SDS-PAGE 2-Dimensional-Sodium Dodecyl Sulfate Polyacrylamide

Gel Electrophoresis

SOD Superoxide Dismutase

Tsa/Ahpc Thiol specific antioxidant/ Alkylhydroxy

xiii

CHAPTER 1

INTRODUCTION

There are now five, and perhaps six, known metabolic pathways by which organisms can grow using CO2 as the sole source of carbon, the Calvin-Benson-

Bassham (CBB) reductive pentose phosphate pathway, the reductive tricarboxylic acid

(RTCA) cycle, the Wood – Ljungdahl acetyl CoA pathway, the 3-hydroxypropionate pathway, and the 3-hydroxypropionate/4-hydroxybutyrate pathway (Berg et al., 2007).

The CBB reductive pentose phosphate pathway is present in most autotrophic organisms, ranging from higher plants to diverse prokaryotes, including photosynthetic and chemolithoautotrophic bacteria. Ribulose 1, 5-bisphosphate (RuBP) carboxylase/oxygenase (RubisCO), the key enzyme of the CBB cycle catalyzes the primary CO2 fixation reaction and is found in some Archaea as well as other photosynthetic organisms (Tabita et al., 2007).

RubisCO is one of the most abundant proteins on earth, as it contributes to roughly 50% of the total soluble protein found in plant leaf tissues and makes up the majority of the protein in some phototrophic microbes. Its abundance may perhaps be attributed to its very poor catalytic efficiency. RubisCO has a turnover number (kcat) of ~

1 5 sec-1, among the lowest for any biological catalyst (Tabita, 1999). As the name indicates, RubisCO is capable of performing both carboxylation and oxygenation of the substrate ribulose 1,5-bisphosphate (Figure 1.1). The oxygenation reaction may be considered a wasteful reaction as it leads to loss of fixed CO2 as a result of oxidative

metabolism of the unique product of the oxygenase reaction, 2-phosphoglycolate.

Moreover, since RubisCO is already a very sluggish enzyme, the fact that O2 competes

with CO2 at the active site makes this enzyme even more inefficient in fixing CO2.

Besides oxygenation, there are other nonproductive side reactions that compete with carboxylation; e.g. isomerization to D-xylulose 1,5-bisphosphate and β- elimination of the

C1 phosphate (Cleland et al., 1998; Pearce and Andrews 2003). It has been suggested that genetic engineering of a more efficient RubisCO enzyme might potentially help meet the ever increasing demand for food and also help to control increasing levels of CO2 in the atmosphere.

2 2- CH2OPO3

HO C H COO- CO2, H2O + COO-

CO2 fixation H C OH

2- 2- 2- CH2OPO3 CH2OPO3 CH2OPO3 C O C O- 3-Phosphoglycerate

H C OH C OH

H C OH H C OH

2- 2- 2- CH2OPO3 CH2OPO3 CH2OPO3 Ribulose 2,3-enediol(ate) COO- 1,5-bisphosphate 2-Phosphoglycolate Oxygenation + COO- O2, H2O H C OH

2- CH2OPO3 3-Phosphoglycerate

Figure 1.1: Carboxylation and oxygenation mechanism of RubisCO.

. 3 On the basis of amino acid sequence differences, the RubisCO family of proteins has been classified into four different groups (Figure 1.2). Form I proteins, the most abundant type, were the first to be discovered and are present in plants, eukaryotic algae, cyanobacteria, and most phototrophic and chemolithoautotrophic bacteria. Form I

RubisCO is comprised of 8 large (~55 kDa) and 8 small (15 kDa) subunits. The remaining three forms of the RubisCO family are oligomers of only the large subunit, with a dimer of large subunits the basic functional unit for all forms of RubisCO. Form II

RubisCOs are present in some proteobacteria and in one eukaryotic type of organism, the dinoflagellates, and perform the typical RubisCO reaction of the CBB cycle. Form III

RubisCOs are so far reported to be present only in archaea (Bult et al., 1996, Galagan et al., 2002 and Klenk et al., 1997). They catalyze typical RuBP carboxylation, which, in some of the organisms has shown to be involved in salvaging intermediates of purine/pyrimidine metabolism (Finn and Tabita 2004, Sato et al. 2007, reviewed in

Tabita et al. 2007) rather than the CBB cycle. All these three forms of RubisCO are able to catalyze RuBP-dependent carboxylation/oxygenation, although with different specificities and under different physiological contexts (Tabita et al., 2007)

The fourth and the most recently discovered group of enzymes are the form IV or

RubisCO-like proteins (RLPs). They are termed RLPs because, despite sharing sequence homology with the bona fide RubisCO counterparts, they are unable to carry out the classical RubisCO reactions. These proteins have thus far been identified in

4 R. rubrum cbbM R. palustris cbbM

B. subtilis RLP M. aeruginaosa RLP G. kaustophilus RLP

R. rubrum RLP R. palustris RLP1

C. tepidum RLP R. palustris RLP2

Figure 1.2: Unrooted Neighbor Joining tree of RubisCO/RLP lineages. The total numbers of sequences considered in each lineage were 35 for I-A, 16 for I-B, 9 for I-C, 22 for I-D, 20 for II, 10 for III-1, 4 for III-2, 20 for IV-NonPhoto, 2 for IV-EnvOnly, 14 for IV-Photo, 16 for IV-DeepYkrW, 12 for IV-YkrW, and 5 for IV-GOS. The width of the arrows is directly proportional to the number of sequences considered for each clade. The scale bar represents a difference of 0.5 substitutions per site. IV-Arc.ful-DSM 4304, Archaeoglobus fulgidus strain DSM4304 (GenBank accession number NP_070416); Met.bur-DSM6242, Methanococcoides burtonii strain DSM6242 (accession number ZP_00563653); Met.hun-JF-1, Methanospirillum hungatei strain JF-1 (accession number YP_503739); Met.the-PT, Methanosaeta thermophila strain PT (accession number ZP_01153096) (Tabita et al., 2007). The names of organisms from which enzymes are mentioned throughout this thesis are in green.

proteobacteria, cyanobacteria, archaea and algae (Hanson and Tabita, 2001, Carré-

Mlouka et al., 2006 and Derelle et al., 2006). RLP was first identified in the photosynthetic, green sulfur bacterium Chlorobium tepidum, which does not have a functional CBB cycle and uses the RTCA pathway for CO2 fixation (Evans et al., 1966).

This type of RubisCO homolog or RLP was also found in the nonphotosynthetic bacterium Bacillus subtilis (Murphy et al. 2002), an organism that does not even have the ability to fix CO2. These observations suggested a function different from CO2

fixation.

RLPs have been further divided into six different subgroups based on sequence

similarities, IV-Photo, IV-Nonphoto, IV-YkrW, IV-Deep YrkW, IV-GOS (Global Ocean

Sequencing) and IV-AMC (Acid Mine Consortium). All the sequences in the RLP lineage

fail to cluster with any bona fide RubisCO sequences from the form I, form II or form III

groups. In addition, all of the RLPs have different substitutions at some of the conserved

amino acid residue positions of the bona fide RubisCOs; most of these residues are

nonconserved. These characteristic features suggest that the RLPs should lack the

ability to perform RuBP-dependent carboxylation, which is what has been found with the

RLPs studied so far. The pattern of differences in active site residue is distinct for the

different RLP lineages (Figure 1.3) suggesting that RLPs of different groups may interact

with different substrates and perform different functions. Later in this dissertation, direct

experimental proof for this hypothesis will be presented. Interestingly, members of the

IV-deep YkrW group, the most diverse group of RLPs, do not seem to follow a common

pattern of active site residue substitutions. Although many RLPs show a lineage-specific

pattern of differences in active site residues, some of the key features of the active site

6 are still conserved. This observation suggests that the function of RLPs from different

lineages may be related to variations in the basic RubisCO/RLP reaction mechanism

(Tabita et al., 2007).

Both RLP and bona fide RubisCOs have been shown to co-exist in some

organisms, e.g. Rhodospirillum rubrum; this organism contains a form II RubisCO as well as a deep-YkrW RLP (Tabita et al., 2007). In addition, Rhodopseudomonas palustris contains both form I and form II RubisCO along with two different RLP types, one from the IV-Photo group and another from the IV-DeepYkrW branch (Larimer et al.,

2004). Finally, the cyanobacterium Microcystis aeruginosa PCC7806 contains a form I

RubisCO and a type IV-YkrW RLP (Carré-Mlouka et al., 2006), while the eukaryotic

alga Ostreococcus tauri contains form I RubisCO and a form IV-DeepYkrW RLP (Derelle

et al., 2006). The presence of both RLP and RubisCO in the same organisms makes

such organisms potentially good systems to study the functional relationship between

these different proteins, while also gleaning some information on the evolution of the

active site of such proteins.

R. rubrum, fixes CO2 via the CBB cycle and the form II RubisCO encoded by its genome is the key CO2 fixing enzyme. A RubisCO disruption strain of R. rubrum (strain

I-19 or cbbM -) was shown to be incapable of growing under photoautotrophic conditions

using hydrogen as the electron acceptor (Falcone and Tabita, 1993). The same

RubisCO disruption strain in a later study was shown to be capable of growth under

photoautotrophic condition, when instead of hydrogen, thiosulfate and sulfide were used

as electron donors (Wang et al., 1993b). Although, photoautotrophic growth of strain I-19

was not strong in this earlier study, these results suggested the possibility of a CO2

fixation pathway different from the CBB cycle.

Lysine carbamylated in RubisCOs Figure 1.3: Conservation of RubisCO active-site residues in RubisCO/RLP family members. Residues are noted in the single-letter IUPAC code. Positions shaded green indicate conservation, yellow indicates a semiconservative substitution and red indicates a nonconservative substitution. C, catalytic residue; R, RuBP binding residue (Tabita et al., 2007). R. palustris RLP1 member of the lineage deepykr.

8 The X-ray structures of RLPs from C. tepidum (Li et al., 2005), R. palustris RLP2

(Tabita et al., 2007) and Geobacillus kaustophilus (Imker et al., 2007) have been solved.

The overall secondary structure of these RLPs is similar to the secondary structure of

bona fide RubisCOs. RLP polypeptides, like RubisCO, are composed of two domains,

an N-terminal α+β domain and a C- terminal (β/α)8-barrel domain. The functional unit is a

homodimer, with the active site, like RubisCO, at the interface of two subunits. There are

two active sites per dimer (Imker et al., 2007; Li et al., 2005).

Many times, functionally related genes are linearly inherited or laterally transferred as functional modules. Analysis of the genomic environments of various

RLPs from different clades of RLP shows that RLPs from different groups are functionally linked to different genes. Like the active site residue pattern, linkage also seems to be conserved within the groups (Figure 1.4). C. tepidum and the other seven green sulfur bacteria of the IV-Photo group have a tightly conserved core of five genes in the RLP region. Two of these are short-chain dehydrogenase/reductase family proteins whereas the other two are conserved hypothetical proteins. B. subtilis RLP, of the IV- ykrW group, which is shown to participate and function in the methionine salvage pathway (MSP), is found in an operon with other genes of the MSP. RLPs of the IV-

NonPhoto group seem to be associated with a conserved gene encoding a surface or secreted protein predicted to be dependent on a type III secretion system for export. The functional significance of the local gene conservation in the groups is currently known only for the IV-YkrW lineage (Tabita et al., 2007).

Figure 1.4: Local gene conservation near genes encoding RLP from the IV-Photo, IV- NonPhoto and IV-YkrW lineages. Gene neighborhoods were visualized using tools at the Integrated Microbial Genomes website (http://img.jgi.doe.gov/cgi-bin/pub/main.cgi). RLP genes are indicated in red. Other open reading frames are colored and identified according to their annotation in the Integrated Microbial Genomes database. R. meliloti; Rhizobium meliloti, B. xenovorans; Burkholderia xenovorans, B. bronchioseptica; Bordetella bronchioseptica, Hrp; type III effector (hypersensitive response protein), EF- Ts; elongation factor-Ts, Bchl; bacteriochlorophyl and SDR, short chain dehydrogenase/reductase (Modified from Tabita et al., 2007).

10 The lack of RubisCO activity has been experimentally confirmed for RLPs from

C. tepidum (Hanson and Tabita 2001), B. subtilis (Ashida et al., 2003), G. kaustophilus

(Imker et al., 2007), R. palustris (S. Romagnoli and F.R. Tabita unpublished results) and

R. rubrum (this study). Experimental evidence regarding the function of RLPs is so far

limited to the work done on RLPs from C. tepidum (putative role in thiosulfate oxidation;

Hanson and Tabita, 2003) and the members of the IV-ykrW group B. subtilis,

Geobacillus kaustophilus and Microcystis aeruginosa (enolase in MSP) (Ashida et al.,

2003, Imker et al., 2007 and Carré-Mlouka et al., 2006).

As stated, prior studies in C. tepidum, specifically with an RLP-disruption strain,

suggests that RLP is involved in sulfur metabolism. The RLP disruption strain shows a

pleiotropic phenotype, with defects in pigmentation, autotrophic growth and thiosulfate

oxidation, whereas this strain was still able to metabolize sulfide. The RLP disruption

strain also shows an imbalance in the cellular levels of low molecular weight thiols. In

addition, the RLP disruption strain produces elevated levels of extracellular elemental

sulfur indicating defect(s) in the oxidation of elemental sulfur. The RLP-gene disruption

also resulted in overexpression of two oxidative stress response genes, tsa (encoding a

thiol specific antioxidant protein) and sod (encoding superoxide dismutase). The precise

reaction catalyzed by C. tepidum RLP is yet to be delineated (Hanson and Tabita, 2001

and 2003).

B. subtilis RLP functions as an enolase in the MSP. It is clear that the RubisCO

substrate (ribulose 1-5, bisphosphate) and the substrate used by this type of RLP (2, 3-

diketo 5-methylthiopentyl 1-phosphate (DK-MTP-1-P)) are structurally similar (Fig. 1. 5).

It is known that the RubisCO CO2 fixation reaction may be divided into three separate steps: (i) enolization, (ii) carboxylation, and (iii) hydrolysis. RubisCO and the IV-Ykrw

11 type RLP perform similar reactions, the enolization partial reaction of the RubisCO reaction scheme, and use structurally similar substrates (Figure 1.5). This suggests that

RLPs from this group are not only structurally similar to RubisCO but they possess some similarity at the functional level as well. Such findings agree with speculation that the

RLP reactions may be some variation of the RubisCO reaction. In support of this idea, It was shown that form II RubisCO from R. rubrum and RLP from cyanobacterium

Microcystis aeruginosa possess in vitro RLP enolase activity and both proteins can functionally complement for the loss of RLP in the RLP disruption strain of B. subtilis

(Ashida et al., 2003; Carré-Mlouka et al., 2006).

CO2 -

Figure 1.5: Comparison of reactions catalyzed by RubisCO and B. subtilis RLP. Both catalyze similar enolase-type reactions and employ structurally analogous substrates. In each instance, a carbamylated lysine catalyzes proton abstraction from the substrate to initialize enolization. DK-MTP 1-P, 2,3-diketo-5-methylthiopentyl-1-phosphate; HK-MTP 1-P, 2-hydroxy-3-keto-5-methylthiopentenyl-1-phosphate (Tabita et al., 2007).

13 One of the phenotypes demonstrated by the C. tepidum RLP disruption strain is over-expression of two genes, tsa and sod, encoding for proteins involved in the oxidative stress response. It was also shown that RLP is not directly involved in overexpression of these genes. Disruption of RLP may elicit a signal which causes over expression of oxidative stress response genes. Identification of transcriptional regulators of tsa and sod gene will be helpful in shedding light on the nature of the signal generated by the RLP disruption, which in turn may give some insight regarding the function performed by this RLP.

RLPs from different groups show group-specific alterations in some of the

RubisCO active site residues and they also show group specific linkage to different genes. Based on these observations it can be hypothesized that RLPs from different groups may perform different functions in their respective organisms and they may not be able to perform functions performed by RLPs from different groups or the bona fide

RubisCO or vice-versa. The ability of the from II RubisCO from R. rubrum to perform enolization of DK-MTP-1-P provides evidence that it may be possible for members of one group to be able to functionally complement members of another group. Systematic studies that involve testing for the ability of different RLPs to perform different functions and complement other RLPs/RubisCO will shed more light on the functional relatedness in the RubisCO super family. These types of studies will also help identify residues critical for the different functions performed by RLPs and RubisCO, which may provide more information regarding the evolution of the RubisCO active site.

Such studies might then eventually provide clues as to determining how all the forms of RubisCO (including RLPs) select between different substrates. In the current study, the function of RLP from R. rubrum was also explored by disrupting the RLP

14 encoding gene. Complementation experiments were then designed that utilize RubisCO and RLP disruption strains from R. rubrum to assist in understanding how alterations in

RubisCO active site residues in different RLPs affect the ability to support MTA- dependent growth.

It was concluded from the current study that genes tsa and sod from C. tepidum, encoding for oxidative stress response proteins, are not under the regulation of PerR or

Fur proteins, which are the conventional regulators in other organisms like B. subtilis, E. coli and S. aureus. Studies with R. rubrum RLP and RubisCO disruption strains show the involvement of both of these proteins in the metabolism of MTA under different physiological conditions. Novel functions for both RLP and RubisCO in their native organism have been described. The pathway of MTA metabolism also seems to be different from the conventionally known MSP. Photoautotrophic growth of a RubisCO disruption strain of R. rubrum was clearly shown and the possibility of whether the ethylmalonyl-CoA pathway of acetate assimilation was partially responsible for photoautotrophic growth was also explored.

CHAPTER 2

CHANGES ELICITED BY DISRUPTION OF THE RUBISCO-LIKE PROTEIN GENE IN

THE GREEN SULFUR BACTERIUM Chlorobium tepidum

INTRODUCTION

Chlorobium tepidum, a gram-negative bacterium of the green-sulfur phylum

(Chlorobi), was originally isolated from high-sulfide hot springs in New Zealand (Wahlund et al., 1991). It is an obligate anaerobic photolithoautotroph, growing optimally at 48°C.

Unlike plants and cyanobacteria, members of Chlorobia perform anoxygenic photosynthesis using hydrogen, or reduced sulfur compounds such as hydrogen sulfide and thiosulfate as electron donors. In addition, instead of using the CBB reductive

pentose phosphate pathway, they perform autotrophic CO2 fixation via the reductive

tricarboxylic acid (RTCA) cycle, using electrons derived from reduced sulfur compounds.

(Evans et al., 1966).

The C. tepidum genome encodes for a homolog of RubisCO, termed the

RubisCO-like protein (RLP). This particular RLP falls under the category of the IV-Photo

clade of RLPs. All the RLPs have nonconservative alterations in some of the conserved

active site residues present in bona fide RubisCOs (Tabita et al., 2007). RLPs of the IV-

16 Photo group are extreme in this regard as they have nonconservative alterations at greater than half the key active site residues (10 out of 19). As expected, based on the differences in the active site residues, C. tepidum RLP does not catalyze the bona fide

RubisCO reactions; i.e. carboxylation or oxygenation of RuBP (Hanson and Tabita

2001).

The C. tepidum RLP disruption strain shows a pleiotropic phenotype. It has a photoautotrophic growth defect, the doubling time of the mutants is three to four-fold reduced compared to the wild type but this does not affect the final cell density. The mutant possesses 20% less of the antenna photopigment, bacteriochlorophyll c as the wild type strain. This mutant also secretes more elemental sulfur outside the cell, indicating some defect in the oxidation of elemental sulfur (Hanson and Tabita 2001).

The RLP disruption strain is also not able to oxidize thiosulfate, whereas the ability to oxidize sulfide is unaffected (Hanson and Tabita 2003).

Both the Tsa [member of thiol specific antioxidant/alkyl hydroperoxide reductase super family of proteins (Tsa/Ahpc)] and superoxide dismutase (SOD) are two major proteins that were shown to hyper-accumulate in extracts prepared from the RLP mutant compared to the wild type strain (Hanson and Tabita 2001). Both these proteins are involved in oxidative stress responses in other organisms. The molecular weight of both proteins is ~25KDa and they co-migrate when analyzed by 1D-SDS as well as 2D-SDS-

PAGE. Moreover, the level of these two proteins is regulated at the level of transcription, as tsa and sod transcripts are 14.2 and 1.7 fold higher in the mutant compared to the wild type (Hanson and Tabita 2003).

17 In Bacillus subtilis the key transcriptional regulator of the inducible peroxide stress

response is a metalloprotein, named PerR, that belongs to a family of ferric uptake

regulator (Fur) proteins initially described to control iron uptake in gram negative bacteria

(Van Vliet et al., 1999). Studies in B. subtilis and some other organisms indicate a link between metalloregulation and the oxidative stress response (Herbig and Helmann,

2002). Typically, PerR and Fur are negative transcriptional regulators but there are reports in some of the organisms where they act as positive transcriptional regulators as well (Herbig and Helmann, 2001).

In B. subtilis and other bacteria PerR is the negative transcriptional regulator of katA (catalase), ahpCF (alkylhydroperoxide reductase), hemAXCDBL (heme

biosynthesis), zosA (zinc uptake) and mrgA (DNA-binding protein) and some other

genes. Genome analysis of C. tepidum showed the presence of homologues of B.

subtilis PerR and Fur proteins (Herbig and Helmann, 2001) in its genome.

The aim of the current study was: (i) to verify the contribution of Tsa and SOD

protein to the hyper accumulated protein band observed in the Ω::RLP strain, this was

studied by making tsa and sod disruption strains; (ii) to study the mechanism underlying

the overexpression of the tsa and sod genes. It is proposed that disruption of RLP elicits

some signal which causes overexpression of tsa and sod genes. Identification of the

regulator(s) may provide information on the nature of the signal, which might be helpful

in deciphering the role of RLP. Involvement of the PerR and Fur proteins in the

regulation of tsa and sod genes was tested by constructing perR and fur gene disruption

strains; (iii) the development of a vector-based complementation system for C. tepidum

to study the ability of various RLPs/RubisCOs to complement for the loss of C. tepidum

RLP. C. tepidum strain Ω::RLP (Hanson and Tabita, 2001) was constructed by insertion

18 of an antibiotic resistant cartridge in the RLP gene. A complete RLP deletion strain

lacking the whole RLP ORF, better suited for complementation studies, was also

constructed.

MATERIALS AND METHODS

Strains and growth conditions. Strains of Chlorobium tepidum used for the

present study were WT2321 (Wahlund and Madigan, 1993) and Ω::RLP (Hanson and

Tabita, 2001). Chlorobium strains were grown in PF-7 medium buffered to pH 6.95 with

10 mM MOPS (3-N-morpholinopropanesulfonicacid)-NaCO3, as described elsewhere

(Mukhopadhyay et al., 1999; Wahlund and Madigan 1993, 1995 and Wahlund et al.,

1991). The growth medium was composed of the following components: EDTA; 0.034

mM, MgSO4 ·7H2O; 0.811 mM, NaCl; 6.84 mM, KH2PO4; 3.67 mM, NH4Cl; 7.48 mM,

ammonium acetate, 6.49 mM, NaHCO3; 24 mM, Na2S2O3 ·5H2O; 12 mM, Na2SeO4;

0.002 mM, Na2WO4 ·2H2O; 0.002 mM, cyanocobalamin; 0.000015 mM, Na2S.9H2O; 2.5 mM and 10 ml of a 100-fold-concentrated mineral solution: nitrilotriacetic acid, 71 µM;

MnCl2.4H20, 4.5 µM; FeCl2.4H20, 6.8 µM; CaCl2, 4.1 µM; CoCl2.6H20, 7.6 µM; ZnCl2, 6.6

µM; CuS04,2.8 µM; Na2MoO4.2H20, 1.9 µM; NiCl2.6H20, 3.8 µM per liter of medium. For growth under autotrophic conditions, ammonium acetate was replaced with an

equivalent amount of NH4Cl. For the selection of transformants and transconjugants,

dilutions of transformation or conjugation mixes were plated on photoheterotrophic

Chlorobium plating (CP) medium plates (Wahlund and Madigan 1995) containing

appropriate antibiotics. Antibiotics used for selection were: spectinomycin 150 µg ml-1, 19 streptomycin 150 µg ml-1, gentamycin 15 µg ml-1, chloramphenicol 15 µg ml-1,

erythromycin 50 µg ml-1 and kanamycin 50 µg ml-1. Plates were incubated in an anaerobic jar with a tube of palladium catalyst to absorb any contaminating oxygen, a

CO2/H2-generating system (5–6% CO2, BBL GasPak system, Becton Dickson

Microbiology Systems, Cockeysville, MD) and a sulfide generating system (0.1 g

thioacetamide and 0.1N HCl). Chlorobium strains were grown at 42 C for antibiotic selection and at 48 C for physiological studies.

The bacterial strains and plasmids used in this study are summarized (Tables 2.1 and 2.2). E. coli strain DH5α was used for cloning all the plasmid constructs and E. coli

strain S17λ pir was used as the donor strain in the conjugation experiments. E. coli

cultures were grown in Luria-Bertani (LB) media containing 1% tryptone, 0.5% yeast

extract, and 1% NaCl (w/v) pH 7.0. Antibiotics used for plasmid selection in E. coli were

ampicilin (100 µg ml-1), kanamycin (50 µg ml-1), gentamycin (15 µg ml-1), erythromycin

(30 µg ml-1) and chloramphenicol (30 µg ml-1). Antibiotics and media components were purchased from Sigma or Fisher.

Molecular biology protocols. Genomic DNA purification, Southern blotting, and polymerase chain reactions (PCR) were carried out by standard protocols

(Mukhopadhyay et al., 1999 and Ausubel et al., 1987). All the genes were amplified from

C. tepidum genomic DNA by performing PCR using Taq DNA polymerase. Plasmid DNA from both E. coli and C. tepidum cells was isolated using a Qiagen plasmid mini prep kit

(Qiagen).

The tsa gene was amplified by PCR using primers 5’-TGTGGCTCTAGAATGGTC

AGGATCGGCG-3’ and 5’-AAGACTCTAGACTACTCCTGGCAGACCGCA-3’ and cloned into the pCR2.1-TOPO vector. Amplification of the sod gene was accomplished using 20 primers 5’-TCTCTAGACACATGGCATATCAGCAACC3’ and 5’ATTCTAGA CGGCAGG

CTTATTTCGCCG-3’ and cloned into pCR-SCRIPT vector (Stratagene). The perR gene was amplified using primers, 5’-GCCCATGAGCAATCGGTATAAAG-3’ and 5’-

AGATCTGCCGGTTACGG TAACGTG- 3’ and was cloned into pCR2.1-TOPO. The fur gene was amplified using primers 5’-ATCGAACATATGGGCAATCAACTCCT-3’ and

5’ATCGCTGGATCCAGCTTGCC-3’ and cloned into pCR2.1-TOPO. For the replacement of corresponding gene with antibiotic resistance gene (see Table 2.1) the antibiotics erythromycin, gentamycin, kanamycin, and chloramphenicol were used for selecting successful transformants in C. tepidum, and to construct knockout mutations in the four genes mentioned above.

Transformation and mutant selection in C. tepidum. All the disruption strains, were constructed by using the transformation procedure described elsewhere (Chung et al., 1998). All knockout strains were constructed by inserting an antibiotic cassette within the coding sequence: the tsa gene was disrupted using a gentamycin resistance cartridge (Tsa::Gm); the sod disruption strain was constructed by inserting a kanamycin gene (Sod::Kn); the perR disruption strain by inserting an erythromycin gene (PerR::Em) and the fur disruption strain by inserting a chloramphenicol gene (Fur::Cm). For the initial selection, dilutions of transformation mixes were plated on photoheterotrophic

Chlorobium plating (CP) medium plates containing appropriate antibiotics. After 7 days colonies from the initial selective plates were picked and re-streaked on selective plates.

After 7 days of growth, single colonies were grown in PF-7 liquid media with the antibiotics. Genomic DNA was prepared from cells from liquid cultures and the genotype of the strains was confirmed by Southern blot analysis. All the disruption strains were constructed in both the wild type as well as the RLP disruption background.

21 Construction of a complete RLP disruption strain RLP∆. The RLP gene

region was amplified via PCR using primers 5’-GGAGTACCATATGGACAAACAGGGGA

-3’ and 5’-GGAACGGATCCCTCAGAAGCATCA-3’ and was cloned into the

pCR2.1TOPO cloning vector. EcoNI and NsiI restriction sites were incorporated on the 5’ and 3’ ends, respectively, of the RLP gene by site directed mutagenesis (Papworth et al., 1996) resulting in the formation of plasmid pRLPNS. Site-directed mutagenesis was performed using primers 5’-CAGTGGCTGACCCTATCAAAGGATGAAT-3’ and its

reverse complement for the incorporation of an EcoNI site on the 5’ end of the RLP gene

and primers 5’-GAAGAAGCAGGACTGAAATGCATCGCTG-3’ and its reverse

complement for the incorporation of an NsiI site at the 3’ end of the RLP gene. The

streptomycin/spectinomycin resistant gene aadA was amplified using primers 5’-

CGTAAGCTGTAATCCTAGTAGTAGAGGATGC-3’ and 5’-TTGAACGAATTGTTATGCA

TTATTTGCC-3’ from plasmid pHP45Ω. The PCR amplified aadA gene was digested with EcoNI and NsiI and was ligated into pRLPNS digested with the same enzymes. This resulted in plasmid pRLP∆, which has the RLP gene replaced by the aadA gene. A linearized pRLP∆ vector was used to transform C. tepidum and to construct a complete

(total) deletion of the RLP gene (strain RLP∆). The RLP deletion was confirmed by

Southern and Western blot analysis.

2D-SDS-PAGE. C. tepidum strains were grown to mid-exponential phase (100 –

120 µg protein ml−1) and harvested by centrifugation at 7500 Χ g for 15 min at room

temperature. Cell pellets were resuspended in 0.5 ml cold lysis buffer (20 mM Tris-Cl,

pH 7.5, 10 mM MgCl2, 1 mM EDTA, 10 mM β-mercaptoethanol, 10% glycerol). Cells

were lysed using a Heat Systems (Farmingdale, New York) model W-385 sonicator

equipped with a microprobe tip. Samples were sonicated for 2 min using a 50% duty 22 cycle, power setting 3.5, while the cell suspension was kept cool in an ice water bath.

Cell debris was removed by centrifugation at 16,000 × g for 10 min at 4 C. Extracts were fractionated by ultracentrifugation at 150,000 × g for 60 min at 4 C. The protein concentration in the soluble fraction was determined with a modified Lowry protocol

(Markwell et al. 1978). Soluble fraction protein samples (100 µg protein) were adjusted to 8 M urea +1% CHAPS + 2 mM tributylphosphine + 0.2% BioLytes 3–10 +0.001% bromophenol blue in a volume of 350 µl. Samples were applied to 17 cm ready strip pH

3–10 IPG strips (Bio-Rad) using the active rehydration program on a Protean IEF cell from (Bio-Rad). Samples were focused for a total of 95,000 Vh (Volt hour) after 15min of prefocusing at using the linear 250 V ramp program from 250 to 10,000 V. SDS-PAGE on pre-cast Protean 8–16% gradient gels (Bio-Rad) was carried out after equilibrating focused IPG strips according to the manufacturer’s instructions. Gels were stained with

Syproruby red (Bio-Rad) and gel images collected using a FluorS MultiImager operated by PDQuest software Version 6.2.1 (Bio-Rad).

Western immunoblots using polyclonal antibodies to C. tepidum RLP.

Antiserum directed against purified C. tepidum RLP (Hanson and Tabita 2001) was used for the Western blot analysis. Proteins resolved by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (Laemmli, 1970) were transferred to polyvinylidene difluoride membranes (Immobilon-P; Millipore, Bedford, MA) according to the manufacturer’s instructions using a BioRad Transblot semi-dry transfer cell (BioRad,

Hercules, CA). Washes and incubations with antibodies were carried out as described elsewhere (Towbin et al., 1979). Antibodies directed against the C. tepidum RLP was used at a dilution of 1:3000 and immunoblots were developed with the Attophos detection reagent according to the manufacturer’s instructions (Amersham,

23 Buckinghamshire, England) and visualized with a Molecular Dynamics Storm 840 imaging system (Molecular Dynamics, Sunnyvale, CA).

Conjugation and selection of transconjugants in C. tepidum. Conjugation was performed by bi-parental matings, the C. tepidum recipient strains were grown overnight in medium PF-7 to the late exponential or early stationary phase (which was equivalent to 100 - 150 µg/ml total cell protein) and then taken into the anaerobic chamber for further manipulations. E. coli S17-λpir strain was used as the donor strain for matings. Overnight cultures of E. coli grown in LB medium with kanamycin were diluted 1:10 in LB (without antibiotic) and incubated at 37 C with shaking for 2 h. Donor cells (1 ml) were then washed, resuspended in LB, and placed inside the anaerobic chamber for mating. Matings were set up by combining recipient cells (1.0 ml) with donor cells (1.0 ml) in an Eppendorf tube and centrifuging them inside the chamber for 4 min at

13,600 X g in a micro-centrifuge. The mating mixture in the pellet was resuspended in 30

µl of CPC medium (CP medium supplemented with 0.05% yeast extract), and the resuspended cells were then spotted onto a CPC medium plate. Control plates containing either recipient cells only or donor cells with empty plasmid without the RLP insert were prepared as described above and included in each conjugation experiment.

The CPC plates containing the mating mixtures were placed in an anaerobic jar and incubated in the light at 42 C in a water bath over night. Sulfide is toxic for E. coli, hence the sulfide-generating system was not employed during this step.

Following mating, jars were taken inside the anaerobic chamber and the cells from each plate were resuspended in 1 ml of CP medium. Dilutions of 10-1 to 10-4 were plated on CP medium plates containing kanamycin. Selection was accomplished in all experiments by incubating the cells in the light at 42 C with the sulfide-generating system

24 until colonies appeared (within 3 to 6 days). Colonies were grown in the PF-7 medium

and cultures were tested for contamination with E. coli by streaking the culture on LB medium plates.

25 Plasmids Relevant characteristics Reference pCR2.1-TOPO Common cloning strain Invitrogen pCR-SCRIPT Common cloning strain Stratagene C. tepidum tsa gene cloned in to pCR2.1- pCTSA This study TOPO pCSOD C. tepidum sod gene cloned in to pcrSCRIPT This study C. tepidum perR gene cloned in to pCR2.1- pCPER This study TOPO pCFUR C. tepidum fur gene cloned in to pCR2.1-TOPO This study Taylor and Rose, pUC4K Kanamycin cassette containing vector 1998 Tichi and Tabita, pUC1318Gm Gentamycin cassette containing vector 2002 Gentamycin gene replaced by chloramphenicol pUC1318Cm This study in pUC1318 Gentamycin gene replaced by erythromycin pUC1318Em This study gene in pUC1318 SmaI digested Gm cassette from pUC1318Gm pCTSA::Gm This study cloned into HpaI site of pCTSA NdeI and BstEII fragment of pCSOD replaced pCSOD::Kn This study by SmaI digested Kn cassette from pUC4K SmaI digested Em cassette from pUC1318Em pCPER::Em This study cloned into NaeI site of pCPER SacI digested Cm cassette from pUC1318Cm pCFUR::Cm This study inserted into SacI site of pCFUR New series of complementation vector for C. Hanson and Tabita pGSS-BBR2 tepidum, see Materials and Methods unpublished results csmB promoter region digested with AseI and pGBC This study NsiI cloned into same sites of pGBSS-BBR2 C. tepidum RLP gene cloned into AseI and pGBC-RLP This study BamHI site of pGBC C. tepidum RLP operon cloned into AseI and pGB-ROP This study BamHI site of pGBC pHP45Ω aadA gene containing vector

pRLPNS See Materials and Mthods This study

pRLP∆ See Materials and Methods This study

Table 2.1: Plasmids used in this study. 26

Bacterial Relevant characteristics Reference strains E. coli

DH5α Common cloning strain Dower et al., 1988 λpir and mobilizing factors on Penfold and Pemberton, S17λpir chromosome 1992 C. tepidum

Wt2321 Wild type strain Wahlund et al., 1991 Hanson and Tabita, Ω::RLP RLP disruption strain 2001 RLP∆ Complete RLP disruption strain This study

Tsa::Gm tsa gene disruption strain This study

Sod::Kn sod gene disruption strain This study

PerR::Em perR gene disruption strain This study

Fur::Cm fur gene disruption strain This study

RLP-Tsa::Gm RLP/tsa double disruption strain This study

RLP-Sod::Kn RLP/sod double disruption strain This study RLP- RLP/perR double disruption strain This study PerR::Em RLP-Fur::Cm RLP/fur double disruption strain This study

Table 2.2: Bacterial strains used in this study.

27 RESULTS

Mutation in sod and tsa genes. Mutations in the tsa and sod genes were made both in the wild type and the RLP disruption (Ω::RLP) background. Both the tsa (Ct

1492) and sod (Ct 1211) disruptions did not result in any noticeable phenotype in the wild type background. Disruption of the sod gene along with the RLP disruption also did not show any apparent difference in the phenotype compared to the wild type. The disruption of the RLP gene resulted in the hyper-accumulation of a protein band (which contains Tsa and SOD proteins; Hanson & Tabita, 2001, 2003) when analyzed by SDS-

PAGE. Disruption of the tsa gene in the Ω::RLP background resulted in the disappearance of the hyper-accumulated protein (Figure 2.1, lane 5).

1 2 3 4 5 6 7

97.4 K.D 66.2 K.D 45.0 K.D

31.0 K.D

Tsa/SOD

21.5 K.D

14.4 K.D

Figure 2.1: Soluble extracts of C. tepidum strains were analyzed by 1D-SDS-PAGE. Tsa-SOD proteins hyper-accumulated in strain Ω::RLP, as seen by 1D SDS-PAGE. This band disappeared in the tsa mutant in the Ω::RLP background. Lanes: 1, wildtype, 2-4, tsa mutant; 5, RLP/tsa mutant; 6, W::RLP and 7, low range standard molecular weight markers (Bio-Rad). Soluble proteins from each strain were prepared as described in Materials and Methods.

Mutation in the perR gene. Disruption of the perR gene resulted in the hyper- accumulation of a protein of the same approximate size as of the Tsa/SOD band when analyzed by 1D-SDS-PAGE (Figure 2.2). Analysis of the total soluble protein by 2D-

SDS-PAGE did not show the over-accumulation of the Tsa/SOD spot (Figure 2.3) which indicates there are some protein(s) other than Tsa and SOD that hyper-accumulate in this mutant strain.

1 2 3 4

97.4 K.D 66.2 K.D

45.0 K.D

31.0 K.D Tsa/SOD band 21.5 K.D

14.4 K.D

Figure 2.2: Soluble extracts of C. tepidum strains analyzed by 1D-SDS-PAGE. The perR mutant strain hyper-accumulated a protein band of approximately the same size as the previously observed Tsa-SOD protein. Lanes: 1, low range standard molecular weight markers; 2, wild type; 3, Ω::RLP; and 4, perR mutant. Soluble extracts from each strain were prepared as described in Materials and Methods.

1 2

Tsa-SOD

3 4

Figure 2.3: Soluble extracts of C. tepidum strains analyzed by 2D-SDS-PAGE. Magnified region showing the absence of the hyper-accumulated Tsa-SOD protein spot in extracts from perR mutant strains. Panel 1, wild type; 2, Ω::RLP; and 3-4, perR mutants.

31 Identification of the hyper-accumulated protein of the PerR mutant strains

by mass spectrometry. Mass spectrometric analysis identified the presence of ferritin

(gene id. Ct1740) and adeninephosphoribosyl transferase (gene id. Ct0293) proteins in the hyper-accumulated band (from the 1D-SDS-PAGE) in the perR mutants.

Inactivation of the fur gene resulted in a pleiotropic phenotype. Inactivation

of the fur gene showed a pigmentation defect as the mutant strains were lighter in color

compared to the wild type at similar biomass densities (Figure 2.4). In addition, the in

vivo absorbance peak for BChl c was 744–745 nm in the Fur disruption strain compared

with 750 nm for WT strain 2321, a 5–6 nm blue shift (Figure 2.5). The in vivo absorption

maximum at 459 nm for the major carotenoid, chlorobactene, was not shifted. The blue

shift was lost when BChl c was extracted into methanol. This observation suggested that

the in vivo blue shift was not because of a structural property of BChl c, but probably

reflected a difference in the degree of aggregation of BChl c.

1 2 3 4 5 6

Figure 2.4: C. tepidum strains grown in PF-7 medium. Tubes: 1, wild type; 2, Ω::RLP; 3- 6, independent isolates of fur mutant strains.

33 A

1.8

1.6 1.4 1.2

1 0.8 0.6 Absorbance 0.4 0.2 0

350 450 550 650 750 850 Wavelength(nm)

B 3

2.5

1.5

Absorbance 1

0.5

350 450 550 650 750 850 Wavelength(nm)

Figure 2.5: A, the absorption spectra of C. tepidum strains showing 4-5 nm blue shift in vivo absorption peak for BChl c. Cells with 40 µg total protein were resuspended in 1 ml 1M sucrose; B; the blue shift was lost when the cells were extracted with 1 ml cold methanol. Green; wild type strain; blue fur gene mutant strain; and red, Ω::RLP strain.

Analysis of the protein profile by 1D-SDS-PAGE showed high levels of two proteins. One stained protein band (Figure 2.6) appeared at the same location as the hyper- accumulated band in the perR mutant (Figure 2.3). 2D-SDS-PAGE analysis showed no difference in the level of Tsa/SOD protein compared to the wild type strain (Figure 2.7).

1 2 3 4 5 6 7 8 9

97.4 K.D 66.2 K.D 45.0 K.D

31.0 K.D A B

21.5 K.D

14.4 K.D

Figure 2.6: Soluble extracts of C. tepidum strains analyzed by 1D-SDS-PAGE showing hyperaccumulated protein bands in the fur mutant strains. A, position of the hyperaccumulated protein band; B, position of the hyperaccumulated Tsa/SOD band. Lanes: 1, low range standard molecular weight marker; 2, extract from wild type strain; 3-8, extracts from independent isolates of fur mutant strain; 9, extract from the Ω::RLP strain

123

Tsa/SOD

Figure 2.7 Soluble extracts of C. tepidum strains analyzed by 2D-SDS-PAGE. Magnified region showing the absence of the over-produced Tsa/Sod protein spot in extracts from the fur gene mutant strain. Panels: 1, wild type strain; 2, Ω:: RLP strain; and 3, fur mutant strain.

36 Construction of complementation vector. The broad host range plasmids

pDSK519 and pGSS33 of the IncQ group can be transferred from E. coli to C. tepidum

by conjugation. These plasmids can be stably maintained by autonomous replication in

C. tepidum below 42 C (Wahlund and Madigan 1995). So far there are no reports in C. tepidum of functional complementation of genes by expression either on the chromosome or on a plasmid. Despite its ability to be maintained in C. tepidum, vector pGSS33 is an inconvenient vector because of the presence of multiple antibiotic resistance markers, no multiple-cloning site, and no immediate screen for insertion of a desired fragment. A newer generation of broad-host range vector with more user-friendly characteristics was thus constructed.

The vector pGSS-BBR2, was constructed by mobilizing the lacZα fragment, multiple cloning site and a kanamycin resistance marker from the pBBR-series plasmids into plasmid pGSS33 (Figure 2.8) (Hanson and Tabita unpublished results).

37 b mo

h p a pGSS-BBR2

Z ' 10849 bp

Figure 2.8: Complementation vector constructed by mobilizing the lacZα fragment, MCS and kanamycin resistance marker from pBBR-series plasmid into the pGSS33 backbone.

Complementation of the RLP disruption phenotype. Plasmid pGBC-RLP was

transferred into the Ω::RLP disruption strain by conjugation. The presence of the plasmid

in C. tepidum was confirmed after preparing plasmid DNA using the Qiagen plasmid miniprep kit. The presence of the RLP protein was checked by Western blot analysis

(Figure 2.9B). The presence of the RLP protein in these experiments confirmed that successful expression of the gene from the complementing plasmid occurred. The resulting transconjugants still showed the pigment defect and hyper-accumulated 38 Tsa/SOD proteins, like the Ω::RLP strain (Figure 2.9). This indicates that even though

the RLP gene was successfully expressed from the plasmid, its addition failed to

functionally complement the RLP disruption phenotype. Complementation was also

attempted by expressing the RLP gene under its own promoter. This was achieved by

transferring plasmid pGB-ROP into the Ω::RLP strain. Transconjugants containing pGB-

ROP also accumulated high levels of the RLP gene but such strains still failed to complement the RLP disruption phenotype (Figure 2.10). Similar complementation experiments were performed using complete RLP deletion strain RLP∆, with the same

lack of functional complementation of the disruption of the RLP gene (result not shown).

A B

1 2 3 4 1 2 3 4 5

97.4 K.D. 66.2 K.D.

45.0 K.D.

31.0 K.D. Tsa/SOD 21.5 K.D.

Figure 2.9: Protein levels in extracts from complemented strains. B, soluble extracts of C. tepidum strains analyzed by 1D-SDS-PAGE. Lanes: 1-2, Ω::RLP/RLP+, 3; Ω::RLP, 4; wild type and 5, Bio-Rad low range standard protein molecular weight markers. A, Western blot analysis using anti-RLP antibodies. Order of lanes same as in B.

A B

1 2 3 4 5 1 2 3 4 5

97.4 K. D. 66.2 K. D.

45.0 K. D.

31.0 K. D.

T sa/SO D

21.5 K. D.

14.4 K. D.

Figure 2.10: A; Analysis of soluble extracts of complemented C. tepidum strains by 1D- SDS-PAGE. Lanes: 1, Bio-Rad low range standard molecular weight markers; 2, wild type strain; 3, Ω::RLP strain; 4-5, Ω::RLP/RLP+ strain. B; Western blot analysis using anti-RLP antibody, showing presence of RLP protein in the Ω::RLP strain complemented with wild type RLP gene. Order of lanes same as in A.

DISCUSSION

In earlier studies it was shown that the level of the Tsa and SOD proteins are 12

and 3 fold higher, respectively, in the RLP disruption strain when compared to the wild

type (Hanson and Tabita 2003). This indicates that Tsa is the major contributor of the

hyper-accumulated Tsa/SOD band. The presence of the hyper-accumulated band in a

sod/RLP double disruption strain agrees with these results as the major contributor, Tsa,

is still intact in the SOD/RLP disruption strain. Disruption of tsa in the Ω::RLP background resulted in the disappearance of the hyper-accumulated band, again agreeing with previous results that indicated that Tsa is the major protein that contributes to the hyperaccumulated protein band observed in extracts from RLP knockout strains.

In many organisms, including Staphylococcus aureus (Horsburgh et al., 2001)., the PerR protein acts as a transcriptional regulator for controlling the expression of genes that encode the iron storage protein, ferritin, as well as genes encoding oxidative stress response proteins, including SOD and Tsa. The hyper-accumulation of the ferritin protein in the C. tepidum perR mutant indicated the potential involvement of the PerR protein as a negative transcriptional regulator of the ferritin gene. Disruption of the fur

gene in C. tepidum resulted in a pleiotropic phenotype, including, a defect in

pigmentation, alteration in the organization of bacteriochlorophyll inside the chlorosomes

and changes in the levels of more than one protein. Pigment defects and a blue shift in

the absorption spectra has been observed in an RLP mutant strain (Hanson and Tabita,

2001) as well as mutations in some of the genes involved in sulfur oxidation (Chan et al.,

2007). This could be a general stress response elicited by physiological changes in the

42 mutant strains. Similar pigmentation defects and blue shift in the bacterial chlorophyll c

absorption spectra can be produced in the wild type by exposing it to light and/or thermal

stress (R Morgan-Kiss and T. E. Hanson, Personal communication). The results with

both perR and fur knockout strains indicated the role of these genes as transcriptional regulators in C. tepidum. However, both PerR and Fur did not seem to be involved in the

regulation of tsa and sod gene expression. These results indicated the involvement of some other regulator(s) involved in the over-expression of the tsa/sod genes caused by

some signal induced by the RLP gene disruption.

Plasmid transfer in C. tepidum via conjugation has already been reported

(Wahlund and Madigan, 1995), but autonomously replicating plasmids have not been exploited for expressing genes for functional complementation in C. tepidum. In the

current study, the RLP gene was shown to be expressed on plasmids pGBC-RLP and

pGB-ROP. However, despite the fact that the RLP protein was shown to be present in

extracts of transconjugants of RLP disruption strains Ω::RLP and RLP∆, such strains did

not exhibit functional complementation as indicated by the inability of such strains to

repair the blue-shift and restore control over Tsp/SOD synthesis. Clearly, from these

results, it is apparent that some or all of the phenotypic changes caused by disruption of

the RLP gene are the result of some indirect effect involving changes elsewhere on the

genome/proteome. Inasmuch as disruption of the gene immediately downstream of the

RLP gene has no effect (Hanson and Tabita, 2001), inactivation of the RLP gene did not

simply cause a polar effect on downstream genes. Further experiments are needed to

resolve the signal transduction mechanism that is affected by expression of the RLP

gene.

CHAPTER 3

ROLE OF RUBISCO AND THE RUBISCO-LIKE PROTEIN IN 5-

METHYLTHIOADENOSINE METABOLISM IN NONSULFUR PURPLE BACTERIA

INTRODUCTION

Rhodosprillum rubrum and Rhodopseudomonas palustris are photosynthetic nonsulfur purple bacteria. They are facultative anaerobes capable of growing autotrophically and heterotrophically under anaerobic conditions using light as the energy source. These organisms can also grow under aerobic heterotrophic conditions using oxygen as terminal electron acceptor. The genome of R. rubrum encodes for a

bona fide form II RubisCO protein as well as a form IV RubisCO (RLP) of the Deep

YkrW group (Tabita et al., 2007). R. palustris possess bona fide form I and form II

RubisCOs and two RLPs, one each of IV-Photo (RLP2) and IV-Deep YkrW group

(RLP1) (Figure 3.1). The function of these RLPs is unknown. Two other photosynthetic

nonsulfur purple bacteria, Rhodobacter sphaeroides and Rhodobacter capsulatus are

also facultative anaerobes capable of both anaerobic photoautotrophic and aerobic

heterotrophic growth. The genomes of both of these organisms encode for bona fide

form I and form II RubisCOs but they do not posses RLP.

Figure 3.1: Minimum evolution phylogenetic tree of the RLP family. R. rubrum and R. palustris, C. tepidum and B. subtilis RLP used in the current study are highlighted in green.

It is possible that different members of this group may be performing different

functions. With the exception of the IV-Deep YkrW lineage, all the other RLP clades

show a lineage-specific pattern of changes in RubisCO active site residues. The IV-

Deep YkrW lineage is the most diverse group of the RubisCO form IV family. The high

level of diversity in the IV-DeepYkrW lineage may eventually lead to further subdivision

of RLP genes in this clade (Tabita et al., 2007). However, our genome analysis shows

that the RLP genes from this lineage seem to be linked to genes encoding for protein

belonging to cupin2 superfamily of proteins (Figure 3.2). It is a very diverse family of

proteins with members identified as dioxygenases, phosphomannose isomerases,

oxalate decarboxylase, etc. (Khuri et al., 2001).

Figure 3.2: Local conservation near genes encoding RLPs of the IV-DeepYkrW lineage in R. rubrum and R. palustris (RLP1), showing conservation of hypothetical proteins of the Cupin2 super family next to the RLPs. Members of the cupin2 super family are indicated by a star. Gene neighborhoods were visualized using tools at the Integrated Microbial Genomes website (http://img.jgi.doe.gov/cgi-bin/pub/main.cgi). RLP genes are indicated in red. Other open reading frames are colored and identified according to their annotation in the Integrated Microbial Genomes database.

46 Methylthioadenosine (MTA) is a by product of spermidine biosynthesis, acyl homoserine lactone and ethylene biosynthesis. In most organisms including plants and humans MTA is converted back to methionine by a Methionine Salvage Pathway (MSP)

(Sekowska et al., 2004). RLPs of the IV-YkrW group catalyze enolization of 2, 3-diketo

5-methylthiopentyl 1-phosphate (DK-MTP 1-P) as part of a Methionine Salvage Pathway

(Figure 3.3). It was also shown that form II RubisCO from R. rubrum possessed in vitro

DK-MTP 1-P enolase activity and the RubisCO gene (cbbM) is able to complement the loss of RLP function in B. subtilis (Ashida et al., 2003). RLP from the cyanobacterium

Microcystis aeruginosa, which has all the genes of MSP present in it genome (Frangeul et al., 2008), also has the ability to catalyze enolization of DK-MTP-1-P and this enzyme is also able to support MTA-dependent growth in the RLP disruption strain of B. subtilis

(Carré-Mlouka et al., 2006). These results point to a common origin between RubisCO and this type of RLP.

There is an oxygen requiring step, the conversion of Dihydroxyketone- methylthiopentene (DHK-MTPene) to Ketomethylthiobutyrate (KMTB) (Figure 3.3) in the

MSP scheme. Thus, any hypothesis regarding the presence of the MSP under anaerobic conditions would require an alternate route bypassing the dioxygenase step. However, at this time there is no experimental evidence of either a functional MSP, or any alternate route bypassing the oxygen requiring step in any organism under anaerobic growth conditions (Sekowska et al., 2004). This is certainly true of R. rubrum and R. palustris as well where, at this time, there is no indication that a functional MSP exists in these organisms. The current study was performed to decipher the role of RLP in R. rubrum and RLP1 and RLP2 of R. palustris and test the ability of RubisCOs present in these organisms to functionally complement for the RLPs and to perform

47 complementation studies to examine the ability of various RubisCOs and RLPs to perform the function of these RLPs.

SAM synthetase O- (Rr0917/3776/ ATP RPA4016) Aminotransferase S O CH3 PPi+Pi (Rr2411/RPA2503) O- H3C O + NH + 3 S Adenine S O O H C - Methionine 3 O + NH3 O S-Adenosinemethionine SAM decarboxylase 2-keto - 4- methylthiobutyrate OH OH (Rr1692) (KMTB) HCOO- H C CO DHK-MTPene 3 2 + + Adenine dioxygenase (mtnZ/ykrZ) H3N S OH O O2 S OH H3C

O OH OH 1-2-dihydroxy-3-keto-5-hydroxy S-Adenosinemethioninamine -methylthiopentene (DHK-MTPene) CH4 Spermidine Putrescine Phosphatase synthetase Pi (mtnX/ykrX) (Rr1691) Spermidine OH

S S Adenine OPO 3-- H3C -- O H3C O 2-hydroxy-3-keto-5-methylthiopentenyl -1-phosphate (HK- MTPenyl-1-P) MTA phosphorylase OH OH Enolase (Rr0361/RPA4821) 5- Methythioadenosine (RrRLP/RubisCO???) O MTA nucleosidase (MTA) S OPO ---- H C 3 3 S OH O H3C Adenine O 2,3- diketo- 5 - methylthiopentyl-1 -phosphate (DK- MTP-1-P) MTR-1-P MTR kinase OH OH Dehydratase -- 5- Methylthioribose H2O S OPO3 O (MTR) O OH H3C ATP -- S OPO -- H C 3 3 ADP MTR-1-P OH isomerase (Rr0360/ OH OH 5- Methylthioribulose-1 RPA4820) 5- Methylthioribose-1 -phosphate (MTRu-1-P) -phosphate (MTR-1-P)

Figure 3.3: Methionine salvage pathway in which the IV-YkrW RLPs, such as the protein from B. subtilis, encoded by the mtnW/ykrW gene, participates in an enolase reaction whereby 2, 3-diketo-5-methylthiopentyl-1-phosphate is converted to 2-hydroxy-3-keto-5- methylthiopentenyl-1-phosphate (highlighted). Homologs of the genes found in R. rubrum and R. palustris CGA 009 are shown in blue and red respectively (adapted from Tabita et al., 2007).

49 MATERIALS AND METHODS

Bacterial strains and growth conditions. R. rubrum strains used in the current

study are Str-2 (wild type, a spontaneous streptomycin (Sm) resistant derivative of strain

S1 (ATCC 11170) and I-19 (cbbM-)(a RubisCO disruption strain) (Falcone and Tabita,

1993). PYE complex medium consisting of 0.3% peptone, 0.3% yeast extract, 10%

Ormerod’s basal salts (Ormerod et al., 1961), and 15 µg of biotin per liter was used for

chemoheterotrophic growth of R. rubrum for conjugation experiments. Ormerod medium

(OM) (Ormerod et al., 1961) containing malate as the carbon source was used for

photoheterotrophic (PH) growth. MTA-dependent growth was achieved with sulfur-

depleted Ormerod’s medium, prepared by replacing the sulfate salts with equimolar

amounts of chloride salts. Antibiotics used for selection of R. rubrum mutants and

transconjugants were kanamycin 50 µg ml-1, gentamycin 10 µg ml-1, tetracycline 36 µg ml-1 and streptomycin 50 µg ml-1.

E. coli strain DH5α was used for cloning all the plasmid constructs; E. coli strain

SM-10 (Simon et al., 1983) was used as the donor strain in the conjugation experiments

and E. coli strain BL21 (DE3) was used for overexpression of the RLP gene for

production of recombinant protein. E. coli cultures were grown in Luria-Bertani (LB)

media containing 1% tryptone, 0.5% yeast extract, and 1% NaCl (w/v). Antibiotics used

for plasmid selection in E. coli were ampicillin at 100 µg ml-1, kanamycin at 50 µg ml-1,

gentamycin at 15 µg ml-1, erythromycin at 30 µg ml-1 and chloramphenicol at 30 µg ml-1.

Antibiotics and media components were purchased from Sigma or Fisher.

50 MTA (Methylthioadenosine) dependent growth of R. rubrum and R.

palustris. Single colonies were used to inoculate culture tubes containing Ormerod’s liquid media under aerobic conditions at 30 C with shaking at 200 RPM (Revolutions Per

Minute), till the mid exponential phase (Optical density660~0.6-0.8). Cells were pelleted

by centrifuging at 12,000 Χ g for 3 min; pellets were washed three times with the sulfur- depleted medium and then resuspended in the same medium. Washed cells were inoculated in sulfur-depleted medium supplemented with MTA. As a negative control in all experiments, cells were also inoculated in sulfur-depleted medium lacking any sulfur source. Anaerobic phototrophic MTA-dependent growth was done by performing the same procedure described above using cells grown chemoheterotrophically inside an anaerobic chamber (Coy labs, Grass Lake, Michigan), maintained at an atmosphere of

2.5 - 3% hydrogen and balance nitrogen. Anaerobic media was prepared under a 100% nitrogen atmosphere and dispensed (10 ml per tube) in 25 ml tubes fitted with butyl rubber stoppers with an aluminum seal crimped over the stopper (Bellco Glass Inc.

Vineland, NJ, USA). Anaerobic cultures were grown in the light at 27 C in a growth

chamber (Environment Growth Chambers, Chagrin Falls, OH, USA). In the experiments

testing for MTA-dependent growth, all the strains were grown in medium without any

sulfur source as a negative control and medium containing methionine (sole sulfur

source) as positive control. The concentrations of MTA and methionine used in the

media were 1 mM. Kanamycin and gentamycin antibiotics which are commercially

available as sulfate salts were not used for MTA dependent growth as they can be a

potential source of sulfate in the medium. All the growth experiments were performed

using three independent isolates and results are representation of at least two

independent experiments.

51 Molecular Biology Protocols. Genomic DNA purification was performed using

the Wizard Genomic DNA purification kit (Promega,). Southern blot analysis and

polymerase chain reactions (PCR) were carried out by standard protocols

(Mukhopadhyay et al., 1999, Ausubel et al., 1987). All the genes were amplified from the genomic DNA of the respective organisms by performing PCR using either Taq DNA polymerase or Pfu DNA polymerase. Plasmid DNA from both E. coli and R. rubrum cells was isolated using a Qiagen plasmid mini prep kit (Qiagen).

Inactivation of the RLP gene. The R. rubrum RLP gene was amplified from genomic DNA by PCR with Pfu DNA polymerase (Invitrogen) using primers 5’-

CGAGGACAGGATCCGCGCCATCGG-3’ and 5’- GCGCCCCTGCAGGCGATCGTCTCC

-3’. The PCR-amplified RLP region was cloned into the pCR-TOPOBLUNT vector

resulting in the plasmid pTBRrRLP. The RLP gene was disrupted after inserting an XmnI and AfeI digested gentamycin cartridge from pUC1318Gm into StuI digested pTBRrRLP.

This resulted in plasmid pTBRLPGm. The disrupted RLP gene was then sub-cloned into suicide vector pSUP202. Plasmid pTBRLPGm was digested with NsiI and the fragment containing the disrupted RLP gene was ligated into PstI digested pSUP202. This resulted in the formation of plasmid pSUP-RLPGm. pSUP-RLPGm was transferred to R. rubrum by conjugation; transconjugants were selected for gentamycin resistance and tetracycline sensitivity. A single RLP mutation was created in the wild type organism. An

RLP and RubisCO double disruption strain was created by disrupting RLP in RubisCO disruption strain I-19 (cbbM-). In the current study, I-19 is referred to as cbbM-. The RLP disruption was confirmed after performing a Southern blot using the RLP gene as a probe.

52 Cloning for complementation studies. Plasmid pRPR was constructed by

cloning the promoter region of the R. rubrum RLP gene into pRK415. The promoter region was amplified by PCR using the following primers: a forward primer incorporating the NdeI site (5’- GGCGTGGATCATATGACGGTGCGCCTGG-3’) and a reverse

incorporating the AseI site (5’- CAGTCTGTCCGTATTAATATGTCTCCCGCGGC-3’).

The PCR product was digested with NdeI and AseI site and cloned into the AseI site of

pRK415. This resulted in plasmid pRPR, which was used for expressing various genes

under direction of the RLP promoter in all complementation studies. A list of primers

used to amplify various RLP and RubisCO genes is given in Table 3.3. All the forward

and reverse primers introduced NdeI and BamHI sites, respectively. All the RLP and

RubisCO genes were digested with NdeI and BamHI restriction enzymes and cloned into

AseI and BamHI digested pRPR.

. Inactivation of the methylthioribose-1-phosphate isomerase (mtrpI) gene.

The region surrounding the mtrpI gene was amplified using the primers 5’- GGGGAA

CATATGTCCGAGGCGTATCGGC-3’ and 5’- GCGACCGCGGATCCGGTCGGGAAACG

AGGCG-3’ and cloned into the pCR-TOPOBLUNT vector. This resulted in plasmid pTB-

MPI. The mtrpI gene was disrupted by inserting an AccI and AfeI digested gentamycin

cartridge into AccI and SrfI digested pTB-MPI. This deleted 245 base pairs of the mtrpI

gene. The disrupted gene was subcloned into suicide vector pSUP202. The disrupted

gene fragment was digested with EcoRI and inserted into the EcoRI site of pSUP202;

this resulted in the formation of plasmid pSUP-MPIGm. Plasmid pSUP-MPIGm was

transferred to wild type R. rubrum by conjugation; transconjugants were selected for

gentamycin resistance. The genotype of the mtrpI disruption strain was confirmed by

Southern blot analysis.

53 Bacterial conjugation and selection of transconjugants. Conjugation was

performed by bi-parental matings. R. rubrum recipient strains were grown 3 - 4 days in

PYE (complex) medium to the late exponential or early stationary phase (Optical

density(OD)660~ 1.2-1.5); the cells were then diluted 1:10 and grown for 1 - 2 days till the mid to late exponential phase (OD660~0.9-1.2). E. coli strain SM-10 was used as the donor strain for the matings. Overnight cultures of E. coli grown in LB medium with appropriate antibiotics were diluted 1:10 in LB (without antibiotic) and incubated at 37 C with shaking at 220 RPM for 2 h. Matings were set up by combining recipient cells (1.0 ml) with donor cells (1.0 ml) in an Eppendorf tube and centrifuging the cells for 4 min at

13,600 Χ g in a micro-centrifuge. This mating mixture pellet was resuspended in 30 µl of

PYE medium and the resuspension was spotted onto a PYE medium plate. Control

plates containing either recipient cells only or donor cells with empty plasmid (pRPR)

without any insert were prepared as described above and included in each conjugation

experiment. The mating PYE plates were incubated in the dark at 30 C overnight.

Following mating, cells from each plate were resuspended in 1 ml of PYE medium. Dilutions of 10-1 to 10-4 were plated on PYE medium plates containing

appropriate antibiotics. The R. rubrum wild type strain is resistant to streptomycin.

Streptomycin was used as a counter selection for E. coli whenever wild type R. rubrum

was the recipient strain. Kanamycin and gentamycin were also used as a counter

selection when the RLP/RubisCO double disruption strain was used as the recipient.

Selection was accomplished in all experiments by incubating plates in the dark at 30 C until colonies appeared (6 to 10 days). Colonies were grown in PYE or OM broth supplemented with appropriate antibiotics and used for further manipulations.

Plasmid Relevant characteristics Reference pCRTOPOBLUNT Common cloning vector Invitrogen R. rubrum RLP gene cloned into pTBRrRLP This study pCRTOPOBLUNT Gentamycin gene inserted into StuI site of pTBRLPGm This study pTBRrRLP R. rubrum mtrpI gene cloned into pTB-MPI This study pCRTOPOBLUNT Gentamycin gene inserted into mtrpI pTB-MPIGm This study gene in pTB-MPI pUC1318Gm Source of Gm cartridge Arsene et al., 1996

pSUP202 Suicide vector Simon et al., 1983 Disrupted RLP gene from pTBRLPGm pSUP-RLPGm This study cloned into PstI site of pSUP202 Disrupted mtrpI gene from pTB-MPIGm pSUP-MPIGm This study cloned into EcoRI site of pSUP202 pRK415 Broad host range vector Keen et al., 1998 Description of construct in materials and pRPR This study methods pRP-RRLP R. rubrum RLP gene cloned into pRPR This study

pRP-CRLP C. tepidum RLP gene cloned into pRPR This study

pRP-BRLP B. subtilis RLP gene cloned into pRPR This study

pRP-PRLP1 R. palustris RLP1 gene cloned into pRPR This study

pRP-PRLP2 R.palustris RLP2 gene cloned into pRPR This study Broad host range plasmid containing R. pRPS-MCS3 Smith and Tabita,2003 rubrum cbbM promoter and cbbR gene R. rubrum cbbM gene cloned into pRPS- pRPS-RrcbbM This study MCS3 R. palustris cbbM gene cloned into pRPS- Satagopan and Tabita, pRPS-RpcbbM MCS3 unpublished R. palustris cbbM I165V mutant cloned Satagopan and Tabita, pRPS-RpI165V into pRPS-MCS3 unpublished R. palustris cbbM I165T mutant cloned Satagopan and Tabita, pRPS-RpI165T into pRPS-MCS3 unpublished Table 3.1: Plasmids used in this study. 55

Strain Relevant characteristics Reference

E. coli

DH5-α Common cloning strain Dower et al., 1988

SM-10 Donor strain in conjugation experiments Simon et al., 1983 High-level expression by IPTG induction of BL21 (DE3) Weiner et al., 1994 T7 RNA polymerase from lacUV5 promoter R. rubrum Falcone and Tabita, Wt (str) Wild type strain 1993 Falcone and Tabita, cbbM- (I19) RubisCO disruption strain 1993 RLP- RLP disruption strain This study

cbbM-/RLP- RubisCO/RLP double disruption strain This study

R. palustris

Wt Wild type strain Lorimer et el., 200 Romagnoli and cbbM-/cbbLS- FormI/FormII double disruption strain Tabita, 200

Table 3.2: Bacterial strains used in this study.

Primer name Primer sequence

RrRLP-NdeF 5’- GGGAGAGTCCACATATGGACCAGTCATC -3’

RrRLP-BamR 5’-AGGGGGGATCCTCCGCGTCTT-3’

RpRLP1-NdeF 5’-TTGGGTTGCA TATGAGCGAG CGGATTATCG-3’

RpRLP1-BamR 5’-CGCGCGGATCCCAACTCGTTCG-3’

RpRLP2-NdeF 5’-GTCATCACATATGACGCCGGACGACATCGC-3’

RpRLP2-BamR 5’-CTACCGCGCTAAGGATCCGGCAATCGC-3’

RrcbbM-NdeF 5’-GCAGGAGATCCATATGGACCAGTC-3’G

RrcbbM-BamR 5’-CGCGCAGGATCCGGAGAACTAC-3’

BsRLP-NdeF 5’-GGTTTTTTCATATGGATGAAAATGAAAGG-3’

BsRLP-NdeF 5’-CGTGATGGATCCGTCAAAATCACAAAT-3’

CtRLP-NdeF 5’-ACCGGATCAACATATGAATGCTGAAG-3’

CtRLP-BamR 5’-GCAGCGGATCCTTTCAGTCCTGCTTC-3’

Table 3.3: Primers used for the complementation studies.

57 Purification of RLP: The R. rubrum RLP gene was amplified using primers, 5’-

GGGAGAGTCCACATATGGACCAGTCATC-3’ and 5’-AGGGGGGATCCTCCGCGTCTT

-3’ containing NdeI and BamHI sites respectively. The RLP gene was cloned into the

NdeI and BamHI site of the pET28a expression vector (Novagen NJ, USA); this resulted

in plasmid pETRrRLP. E. coli BL-21(DE3) cells containing the transformed pETRrRLP

vector was grown in 100 ml LB overnight. The culture was diluted 1:10 into 2.8 l broad

bottom flasks containing 1 l of LB media and grown to an OD of 0.4 - 0.6 at 37°C with 600

shaking at 120 rpm. Gene expression was induced by adding IPTG to the culture. The

culture was grown for 16 h at room temperature with shaking at 120 rpm.

Prior to column chromatography, cells were resuspended in wash buffer (50 mM

NaH2PO4, 300 mM NaCl, pH 8.0) supplemented with 10 mM phenylmethylsulfonyl fluoride (PMSF) and 50 µg/ml deoxyribonucleic acid I (DNase I) and were disrupted using a pressurized French pressure cell (at 110,000 kPa). Lysed cells were then centrifuged at 16,000 g at 4 C for 15 min. The supernatant was further ultracentrifuged at

18,000 g for I h at 4 C. Ni-NTA slurry was added to the supernatant and incubated at 4

C for 1 h with gentle rocking. The supernatant-slurry (Qiagen) mixture was then poured into a 2.5 cm diameter column and allowed to settle for 15 min. The column was then washed with wash buffer containing 20 mM imidazole to elute unbound proteins. The

RLP protein was eluted with the elution buffer ( wash buffer containing 250 mM

Imidazole); 1 ml fractions were collected. Fractions were analyzed by SDS-PAGE.

Fractions containing RLP were pooled and dialyzed against wash buffer without imidazole in order to remove imidazole. Further column chromatography was performed using a Bio-Rad BioLogic HR Workstation. Dialyzed samples were filtered using 0.45 µm filters and loaded onto a Superose 6 gel filtration column (Pharmacia) equilibrated with

58 wash buffer. Samples were eluted with elution buffer (50 mM NaH2PO4, 300 mM NaCl,

pH 8.0). Fractions were analyzed by SDS-PAGE and fractions containing the RLP were

pooled and loaded onto an UnoQ anion exchange column (Bio-Rad) equilibrated with

wash buffer. Samples were eluted using a gradient of 0-1 M NaCl in the buffer. Fractions

were analyzed by SDS-PAGE for the presence of RLP.

Synthesis of 2, 3 ketohexane 1-phosphate (DK-H 1-P), a substrate analog of

DK-MTP 1-P. DK-H 1-P was synthesized in collaboration with Dr. Peng George Wang’s

group at The Department of Biochemistry, The Ohio State University (R. S. Perali and P.

G. Wang, personal communication). DK-H 1-P was synthesized as described elsewhere

(Zhang et al., 2004).

Spectrophotometric assay using 2, 3 diketohexane 1-phosphate (DK-H 1-P).

Spectrophotometric assays were performed with a Perkin-Elmer Lamda 14 UV-vis-

-1 -1 spectrophotometer at 278 nm (ε278 = 2000 M cm for the enol tautomer of HK-H 1-P).

Reaction mixtures (200 µL) contained 0.1-5 mM DK-H 1-P, 5 mM MgCl2, 25 mM

NaHCO3, and 25 mM potassium HEPES (pH 7.5). Enzyme concentrations ranged from

0.5 to 2 µM.

59 RESULTS

Correlation between the presence of RLP and the presence of a functional

Methionine salvage pathway (MSP). Four nonsulfur purple bacteria: R. rubrum, R. palustris, Rhodobacter sphaeroides and Rhodobacter capsulatus were tested for the presence of MSP by checking their ability to utilize methylthioadenosine (MTA, an intermediate in the MSP) as the sole sulfur source. They were grown in Ormerod’s media containing MTA as the sole sulfur source. R. rubrum and R. palustris were able to grow by metabolizing MTA whereas the other two organisms, R. capsulatus and R. sphaeroides, lacked the ability to grow and metabolize MTA when it was supplied as sole sulfur source (Figure 3.4). This result showed a correlation between the presence of

RLP and a functional MSP, as the genomes of R. rubrum and R. palustris encode for

RLP genes, whereas the other two organisms lack RLP genes. All the four organisms were however able to grow on Ormerod’s media containing methionine as the only sulfur source (Figure 3.4), which shows R. capsulatus and R. sphaeroides do not lack the ability to metabolize methionine. Thus, it is the pathway for the conversion of MTA to methionine which is missing in these organisms.

The genomes of both R. rubrum and R. palustris encode for homologs of most of the genes involved in the MSP. Genome analysis further showed that the MSP in R. rubrum can be clearly traced up till the formation of 5- methylthioribulose 1-phosphate

(MTRu 1-P) (Figure 3.3). Except for RLP, there are no clear homologs of proteins involved in catalyzing later steps in the pathway, raising the possibility of some novel proteins that catalyze these remaining steps or perhaps the presence of a novel alternative route for metabolizing MTA.

0.8

0.6

0.4

OD 660 nmm nmm 660 OD

0.2

0 0 25 50 75 100 125 150 Time (hours)

0.8 m 0.6

0.4 OD 660 nmOD

0.2

0 0 25 50 75 100 125 150 Time (hours)

Figure 3.4. Growth of four purple nonsulfur bacteria on methionine (A) and MTA (B) as the sole sulfur source under aerobic conditions. R. rubrum (◆), R. palustris (•), R. capsulatus (•) and R. sphaeroides (•).

61 RLP and RubisCO disruption strains of R. rubrum. The RLP gene was disrupted after insertion of a gentamycin gene cartridge within the coding sequence. The genotype was confirmed after Southern blot analysis (Figure 3.5). The gentamycin insertion lead to an increase in the band corresponding to the RLP gene. The RLP gene was disrupted both in the wild type as well as in the RubisCO disruption (cbbM-) background. Single RLP disruption and RLP/RubisCO double disruption strains did not show any apparent phenotypic difference when grown both under phototrophic and chemotrophic conditions, when compared to their respective parent strains.

62 A B

1 2 3 4 5 2 3 4 5

5000 bp 4000 bp 3000 bp

2000 bp 1600 bp

1000 bp

Figure 3.5. Southern blot analysis showing disruption of the RLP gene. A: agarose gel, Lanes: DNA containing (1); Invitrogen 1Kb+ standard, (2); cbbM-/RLP- strain, (3); RLP- strain, (4); cbbM- strain, and (5); wild type strain. B: Southern blot, wild type RLP gene was used as the probe. The wild type RLP gene (1622 bp), the disrupted RLP gene containing the insertion of the gentamycin cartridge, caused a shift in the band corresponding to the RLP gene (2699 bp). Bp, base pairs.

63 Role of RLP and RubisCO in MTA-dependent growth. The role of R. rubrum

RLP was investigated in the RLP disruption strain by testing for its ability to grow using

MTA as the sole sulfur source. The RLP disruption strain was apparently incapable of

using MTA as the sulfur source under aerobic growth conditions (Figure 3.6). This result

indicated that RLP was involved in metabolizing MTA. The RubisCO disruption strain

(cbbM-) was able to grow using MTA as the sole sulfur source under this condition. As

expected, the RubisCO/RLP double disruption strain (cbbM-/RLP-) was unable to grow using MTA as the sole sulfur source under aerobic conditions (Figure 3.6).

64 1.5

OD660 nm OD660 0.5

0 0 20 40 60 80 100 120 140 160 Time (hours)

Figure 3.6. Growth of R. rubrum strains under aerobic conditions on Ormerod’s - - medium containing MTA as the sole sulfur source. Wild type (◆), RLP (∆), cbbM (□) and cbbM-/RLP- (X). Malate was used as the carbon source.

65 Wild type strains of R. rubrum can also utilize MTA as the sole sulfur source under anaerobic conditions. This is surprising because a functional Methionine salvage pathway requires oxygen(Figure 3.7). Further analysis showed that the R. rubrum RLP-

strain, which cannot metabolize MTA under aerobic growth conditions, as demonstrated

by the inability to grow using MTA as sole sulfur source (Figure 3.6), has the ability to

metabolize it under anaerobic growth conditions (Figure 3.7). It was further observed

that the R. rubrum cbbM- strain was barely able to grow with MTA as sole sulfur source under anaerobic conditions, indicating the potential involvement of RubisCO (instead of

RLP) in MTA metabolism under these growth conditions. Strain cbbM-/RLP-, as expected, was unable to grow using MTA as the sole sulfur source under anaerobic conditions (Figure 3.7). Because of the inability to metabolize MTA both under aerobic as well as anaerobic conditions, the cbbM-/RLP- strain was used for further

complementation studies.

66 2.5 2 1.5 1

O D660 nm 0.5 0 0 48 96 144 192 240 288 336 Time (hours)

Figure 3.7. Growth of R. rubrum strains under anaerobic conditions on Ormerod’s medium containing MTA as the sole sulfur source. Wild type (), RLP- (∆), cbbM- (□) and cbbM-/RLP- (X).

Complementation of R. rubrum cbbM-/RLP- strain under aerobic growth conditions. Strain cbbM-/RLP- was used for all complementation studies. Different RLPs

and RubisCOs were tested for their ability to support MTA-dependent growth in the

cbbM-/RLP- strain. Different genes were expressed on plasmid pRPR, a derivative of pRK415. Plasmid pRPR was constructed by cloning the promoter region of the RLP gene from R. rubrum into pRK415 (Table 3.1); this ensured similar levels of expression for all the different genes under direction of the R. rubrum RLP promoter. Genes 67 encoding for form I RubisCO from R. palustris, form II RubisCO from R. rubrum and

RLPs from R. rubrum (RrRLP), C. tepidum (CtRLP), B. subtilis (BsRLP) and R. palustris

RLP1 (RpRLP1) were tested for their ability to support MTA-dependent growth under aerobic growth conditions. As expected, RLP from R. rubrum, when expressed on the plasmid, was able to complement MTA-dependent growth of the cbbM-/RLP- strain

(Figure 3.8 A). B. subtilis RLP was able to partially rescue the MTA-dependent growth phenotype. All other proteins, C. tepidum RLP, R. palustris RLP1, R. palustris form I and form II proteins (e.g., CbbLS and CbbM respectively) were not able to support MTA- dependent growth.

1 A 1 B 0.9 0.9 0.8 0.8 0.7 0.7

0.6 m 0.6 0.5 0.5 0.4 0.4 OD 660n OD OD 660 nm 660 OD 0.3 0.3 0.2 0.2 0.1 0.1 0 0 0 40 80 120 160 200 240 280 0 40 80 120 160 200 240 280 Time (hours) Time (Hours)

1 C 1 0.9 0.9 D 0.8 0.8 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 OD nm 660 0.3 OD 660nm 0.3 0.2 0.2 0.1 0.1 0 0 0 40 80 120 160 200 240 280 0 40 80 120 160 200 240 280 Time (hours) Time (hours)

1 E 1 0.9 0.9 F 0.8 0.8 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 OD 660 nm OD nm 660 0.3 0.3 0.2 0.2 0.1 0.1 0 0 0 40 80 120 160 200 240 280 0 40 80 120 160 200 240 280 Time (hours) Time (hours)

Figure 3.8. Growth of the R. rubrum cbbM-/RLP- strain complemented with different RLP and RubisCO genes under aerobic conditions with plasmids containing the following genes: (A), R. rubrum RLP, B, B. subtilis RLP; (C), C. tepidum RLP; (D), R. palustris RLP1; (E), R. palustris cbbM; (F), R. palustris cbbLS. Growth on medium containing methionine (■), MTA (□) and no sulfur source (X).

Complementation of the cbbM-/RLP- disruption strain under anaerobic growth conditions. The R. rubrum cbbM gene when expressed on plasmid (pRPS-

RrcbbM) was able to rescue the MTA-dependent growth phenotype of the cbbM-/RLP-

strain under anaerobic growth conditions (Figure 3.9). The ability of form II RubisCO

from R. palustris to support MTA-dependent growth was also tested by expressing the

R. palustris cbbM gene using plasmid pRPS-MCSRpcbbM. The wild-type form II

RubisCO protein from R. palustris was also able to support MTA-dependent growth

under anaerobic conditions (Figure 3.10).

From these results it is clear that RubisCO is interacting with a different substrate

in MTA metabolism and catalyzes a reaction different from the typical

carboxylation/oxygenation of RuBP as a part of the CBB cycle (Bassham et al., 1950).

Residue 165 of the R. palustris CbbM protein is critical for RubisCO activity. The R.

palustris cbbM gene was mutated by site-directed mutagenesis in order to alter residue

isoleucine at residue 165 into either threonine or valine, creating the I165T and I165V

mutant proteins, respectively. The I165V enzyme was able to support photoautotrophic

growth of R. capsulatus strain SBI/II- (Satagopan and Tabita, unpublished results). R.

capsulatus SBI/II- has both of its endogenous RubisCO genes knocked out (Paoli et al.

1998), whereas the I165T protein was not able to support photoautotrophic growth

(Satagopan and Tabita, unpublished results). By contrast, the mutant I165V and I165T

R. palustris CbbM (form II) proteins, expressed from plasmids pRPS-RpI165V and

pRPS-RpI165T were able to support MTA-dependent growth in the cbbM-/RLP- strain under anaerobic growth conditions (Figure 3.10; Table 3.4).

70 2.5

1.5

1 660OD nm

0.5

0 0 100 200 300 400

Time (hours)

Figure 3.9. Growth of R. rubrum strain cbbM-/RLP- under anaerobic conditions on Ormerod’s medium containing MTA as the sole sulfur source. Strain cbbM-/RLP- with no - - plasmid (◊); strain cbbM RLP containing plasmid RrcbbM (◆).

2 1.8 1.6 1.4 1.2 1 0.8 OD 660 OD 660 nm 0.6 0.4 0.2 0 0 100 200 300 400 Time (hours)

Figure 3.10. Growth of R. rubrum cbbM-/RLP- strain under anaerobic conditions on Ormerod’s medium containing MTA as the sole sulfur source using the following plasmids: RpcbbM (●); RpcbbMI165T (∆) and RpcbbMI165V(□); no plasmid (X).

a Photoheterotrophic and Photoautotrophic Strain MTA dependent growth in growth in R. rubrum cbbM-/RLP- R. capsulatus SBI/II-

Wild type + +

I165V + +

I65T + -

Table 3.4: Summary of the ability of R. palustris strains to support MTA-dependent growth in R. rubrum and PA growth in R. capsulatus strain SBI/II-. a Results from Satagopan and Tabita (unpublished results).

MTA-dependent growth of R. palustris strains. The ability of the wild type R. palustris strain to grow using MTA as the sole sulfur source was tested. The wild type strain of R. palustris has the ability to use MTA as the sole sulfur source both under aerobic and anaerobic conditions. As mentioned earlier, the R. palustris genome encodes for two RLPs (RLP1 and RLP2) and two bona fide RubisCOs. Their involvement in the metabolism of MTA was further investigated by testing mutants available that contain disruptions of various genes encoding for these proteins. R. plaustris RLP1-, RLP2-, formI-/formII- and formI-/formII-/RLP2- strains (Romagnoli and

Tabita, unpublished results) were tested for their ability to metabolize MTA. It was

observed that all of these strains were able to grow using MTA as the sole sulfur source

under aerobic as well as anaerobic conditions (Table 3.4). These results show that in the

73 absence of either RLP, one of the two RLPs or one of the RubisCOs is able to support growth on medium with MTA as the sole sulfur source, both under aerobic and anaerobic conditions. Unlike R. rubrum where the cbbM- strain is unable to use MTA as the sole sulfur source under anaerobic conditions, the R. palustris formI-/formII- strain was able to grow using MTA as the sulfur source under anaerobic conditions, suggesting that one of the RLPs may be involved in utilizing MTA.

1.5 m

1 OD 660n

0.5

0 0 50 100 150 200 250 Time (Hours)

Figure 3.11. Growth of R. palustris wild type strain on Ormerod’s medium containing MTA as the sole sulfur source. Growth under anaerobic (▲) and aerobic conditions (■).

Strains Aerobic Growth Anaerobic Growth

Wt + +

FormI-/FormII- + +

RLP1- + +

RLP2- + +

FormI-/FormII-/RLP2- + +

Table 3.5. Table summarizing growth of R. palustris strains on Ormerod’s media with MTA as sole sulfur source, under aerobic and anaerobic growth conditions.

Disruption of the R. rubrum methylthioribose-1-phosphate isomerase

(mtrpI) gene. The mtrpI gene was disrupted by insertion of a gentamycin cartridge into

the open reading frame. The mutants were tested for the ability to metabolize MTA.

Mutants in the mtrpI gene were not able to grow on the medium containing MTA as the

sole sulfur source both under aerobic as well as anaerobic conditions (Figure 3.12 A and

B). This indicated the requirement of this protein encoded by the mtrpI gene for MTA

metabolism under both aerobic and anaerobic growth conditions.

76 A

2 1.8 1.6 1.4 1.2 1 0.8 OD 660 nm 0.6 0.4 0.2 0 0 50 100 150 200 250 Time (Hours)

Figure 3.12: Growth of R. rubrum wild type (■) and mtrpI- (□) strain under aerobic (A) and anaerobic (B) conditions. 77

In vitro activity of recombinant R. rubrum RLP. Recombinant R. rubrum RLP was purified after expressing the RLP gene in E. coli BL21 (DE3) (Figure 3.13). The RLP protein lacked the ability to catalyze RuBP-dependent carboxylation or oxygenation

(data not shown). The natural substrate for the B. subtilis RLP (DK-MTP-1-P) is unstable; thus R. rubrum RLP was tested for the ability to catalyze the enolase reaction using DK-H 1-P, an analog of the substrate for the enolase phosphatase from Klebsiella oxytoca (Zhang et al., 2004). The R. rubrum RLP did not catalyze enolization of the substrate analog, while the B. subtilis protein did (Table 3.6).

Enzyme Specific Activity (µmol.min-1.mg-1)

B. sutilis RLP 5.5

R. rubrum RLP < 0.1

Table 3.6: Specific activity of B. subtilis and R. rubrum RLPs as DK-H 1-P enolase.

78 A * * * 97.4 K. D. 66.2 K. D. 45.0 K. D.

31.0 K. D.

21.5 K. D.

14.4 K. D.

B * * C 97.4 K. D. 66.2 K. D. 45.0 K. D.

31.0 K. D.

21.5 K. D.

14.4 K. D.

Figure 3.13: Coomassie-stained SDS-PAGE of samples from the R. rubrum RLP purification. The R. rubrum RLP gene was expressed in E. coli BL21 strain and samples obtained from: the Ni column (A); Superose-12 gel filtration column (B); and UnoQ anion exchange chromatography (C). Fractions indicated by * were pooled and used for the subsequent step.

79 As already stated the natural substrate for the B. subtilis RLP (DK-MTP-1-P) is unstable. Thus, in order to assay for the enolase activity the substrate is made fresh by converting MTRu-1-P into of DK-MTP-1-P using the enzyme MTRu-1-P dehydratase from B. subtilis. Interestingly, R. rubrum RLP does not have the ability to perform enolization of DK-

MTP-1-P. In fact it was found as a result of NMR studies that RLP from R. rubrum actually used MTRu-1-P as substrate and catalyzed an isomerase reaction, converting MTRu-1-P into 1-thiomethyl-D-xylulose-5-phosphate and 1-thiomethyl-D-ribulose-5-phosphate. The two products were found to be produced in the ratio of 3:1 (Imker and Gerlt, personal communication) (Figure 3.14).

80 A

S H C H 3 O OH O O OH H H -2 -2 -2 S OPO3 S OPO S OPO3 H -2 H2O 3 OPO3 H3C H3C H3C HO OH OH O RLP O MtnS MtnY MTR-1-P MTRu-1-P DK-MTP-1-P HK-MTPenyl-1-P

B 81 OH O O O OH - - OH OH S - - S - - H C OPO3 S OPO H C OPO3 3 RrRLP H3C 3 + 3 OH OH Methylthioribulose 1-thiomethyl-D-xylulose 1-thiomethyl-D-ribulose -1-Phosphate -5-phosphate -5-phosphate 3 : 1

81 Figure 3.14: A: Enolase reaction catalyzed by RLPs from IV-YkrW group (Ashida et al., 2003). B: Proposed reaction catalyzed by R. rubrum RLP (Imker and Gerlt, personal communication).

DISCUSSION

Both R. rubrum and R. palustris wild type strains were shown in the current study to be able to grow using MTA as the sole sulfur source, both under aerobic and anaerobic growth conditions. In earlier studies it was shown that RLPs of the form IV-

YkrW group catalyze enolization of DK-MTP-1-P as part of the MSP. It was also shown that form II RubisCO from R. rubrum, the key CO2 fixing enzyme in the CBB cycle, also

possesses the ability to perform enolization of DK-MTP-1-P, both in vitro as well as in

vivo by virtue of its ability to complement 5-methylythioadenosine (MTA)-dependent

growth of a ykrW/mtnW mutant of B. subtilis (Ashida et al., 2003). Probably due to the oxygen requirement of one of the key enzymatic steps, the MSP, so far has been shown to be functional only under aerobic growth conditions. The presence or significance of the MSP has not been previously investigated in any of the organisms considered in the current study. MTA, a byproduct of spermidine biosynthesis/ethylene biosynthesis/N- acylhomoserine lactone synthesis (and probably other uncharacterized pathways), is a key intermediate of the MSP. In this pathway, MTA is converted back to methionine in a variety of organisms, including plants and humans (Sekowska et al., 2004) (Figure 3.3).

In some bacteria, including E. coli and Salmonella enterica, sulfate is reduced to sulfide and further assimilated into the amino acid cysteine. Cysteine provides the sulfur atom for other sulfur-bearing molecules in the cell, including methionine. These organisms cannot use methionine as a sole source of sulfur due to the lack of a pathway to convert methionine to cysteine (Seiflein and Lawrence, 2001). This constraint is not shared by many other organisms, including the nonsulfur purple bacteria of the current

83 study, which can grow using either cysteine or methionine as the sole source of sulfur.

The ability of these organisms to grow using methionine, and in turn employ MTA as the

sole sulfur source via a functional MSP, forms the basis of the-MTA dependent growth

experiments conducted in the current study.

The genomes of both R. rubrum and R. palustris encode for a majority of the

enzymes required for a functional MSP. Most importantly, both organisms contain at

least one RLP gene. Based on their ability to grow using MTA as sole sulfur source, R.

rubrum and R. palustris possess a functional MSP. On the other hand, as expected, both

R. capsulatus and R. sphaeroides, which do not possess RLP and lack genes encoding

for most of the MSP proteins, lack a functional MSP as these organisms cannot use

MTA as a sulfur source. Certainly, the involvement of RLP in organisms that contain a

functional MSP has already been demonstrated for organisms that belong to the IV-

YkrW group. Thus, the involvement of the RLP from R. rubrum in this pathway was not

completely unexpected. However, based on bioinformatics and structural analyses

(Tabita et al. 2007) and previous studies with C. tepidum RLP (Hanson and Tabita, 2001 and 2003), it seemed apparent that RLPs from different groups might be performing different functions. Indeed, as will be discussed in greater detail later, despite using the same basic pathway, RLPs of the IV-YkrW and IV-DeepYkrW clades apparently perform different functions. Moreover, although R. rubrum RubisCO has the ability to catalyze enolization of DK-MTP-1-P (Ashida et al. 2003), the inability of the RLP disruption strain to exhibit aerobic MTA-dependent growth (by virtue of the presence of RubisCO and an intact cbbM gene) was not unexpected. This is because of the low expression of

RubisCO under aerobic growth conditions and the fact that RubisCO from R. rubrum had

84 previously been shown to be oxidatively inactivated and subsequently degraded under aerobic growth conditions (Cook and Tabita 1988; Cook et al., 1988).

It was observed that the R. palustris double RubisCO deletion strain (formI-/II-) was able to metabolize MTA under anaerobic growth conditions. This is different from results with the R. rubrum RubisCO disruption strain which is unable to metabolize MTA under similar growth conditions, suggesting that in R. rubrum, RLP may perform a function different from its ability to catalyze enolization of DK-MTP-1-P under anaerobic growth conditions. Certainly, there may be other reasons for this difference in the two organisms, including the fact that R. palustris contains two RLPs, one (RLP1) related to the R. rubrum protein and one (RLP2) more related to the C. tepidum RLP, which definitely performs a function different from the YkrW group (Hanson and Tabita 2001;

2003). In addition, the differences observed with respect to anaerobic growth on MTA might be related to different levels of RLP gene expression in the two organisms.

Genes encoding RLPs from C. tepidum and R. palustris (RLP1), when expressed on a broad-host range plasmid, were unable to support aerobic MTA-dependent growth of the cbbM-/RLP- strain of R. rubrum. C. tepidum RLP falls under the IV-Photo group of

RLPs, and as previously noted, based on sequence and structural analyses, RLPs of the

IV-Photo group are different from the RLPs from both the IV-YkrW and IV-DeepYkrW group. With respect to conserved RubisCO active site residues, the pattern of the differences is certainly distinct for different groups of RLPs. RLPs from IV-YkrW, IV-

DeepYkrW and IV-Photo group differ substantially from each other in the active site residue pattern, which, along with other differences, might partially explain their inability to complement each other. The inability of form I (CbbLS) and form II (CbbM) RubisCO from R. palustris to support MTA-dependent growth of the cbbM-/RLP- strain of R.

85 rubrum under aerobic conditions may be attributed to the poor expression of these genes under aerobic conditions. Since the R. rubrum RubisCO protein is degraded under aerobic conditions, it is also possible that the R. palustris RubisCO proteins also

become degraded via a similar mechanism in a R. rubrum background.

The presence of a MSP under anaerobic growth conditions has long been

speculated but so far there has been no documented report of this pathway being used

under these conditions. Because of the oxygen requiring dioxygenase step of the MSP,

as constituted, it is not possible for the pathway to function without oxygen. This

suggests that an alternate route to bypass the dioxygenase step must be present in

order for there to be a functional MSP or alternative pathway. In the current study it is

shown that a functional MSP is present in both R. rubrum and R. palustris under

anaerobic growth conditions. Thus, it is very likely that there is some alternate route to

avoid the oxygen-requiring step in these organisms. Moreover, there are no homologous

genes beyond the RLP-catalyzed step found within the genomes of these two

organisms, suggesting that the aerobic part of this pathway might be different as well.

The R. rubrum and R. palustris cbbM genes were able to support MTA-

dependent growth of the cbbM-/RLP- strain when expressed on a broad-host range plasmid under anaerobic growth conditions. In addition, the fact that these genes could not support photoautotrophic (PA) growth in the same strain may be explained by the fact that it had been previously shown that the whole cbb operon was required to complement the ability of the R. rubrum cbbM knockout strain to support PA growth

(Falcone and Tabita 1993). Most importantly, from the studies thus far completed it is apparent that the RubisCO enzyme catalyzed two separate reactions in R. rubrum and this enzyme plays a major role in two separate and distinct pathways, the CBB CO2

86 assimilatory cycle and an MTA scheme (MSP) to restore methionine. Clearly, RubisCO

has the capacity to interact with two different (yet structurally analogous) substrates, as

previously demonstrated (Ashida et al. 2003). However, the current studies with R.

rubrum indicated that RubisCO was particularly important for anaerobic MTA-dependent growth but RubisCO did not seem to be required for aerobic growth on MTA.

Interestingly, the opposite situation was found with respect to RLP, namely that RLP was required for aerobic MTA-dependent growth but not anaerobic MTA-dependent growth.

Such in vivo studies point to the fact that RubisCO and RLP play distinct physiological

roles relative to MTA metabolism depending on whether the organism is grown in the

presence or absence of oxygen. Clearly, such studies also impinge on the plasticity and

flexibility of RubisCO’s active site to function in a physiologically relevant fashion in two

separate and important pathways. Based on knowledge of the structure and reaction

mechanism employed by RubisCO, it may be assumed that both substrates might

interact at the same active site (Li et al 2005; Tabita et al. 2007). However, it is very

likely that active site residues critical for reacting with RuBP are not so important for

interacting with the second substrate. Residue Ile-165 of R. palustris form II RubisCO

(I164 in R. rubrum RubisCO; Chene et al., 1997) is one such example. An I165T

substitution in the R. palustris form II RubisCO was not able to support PA growth of R.

capsulatus SBI/II- (a RubisCO deletion strain) (Satagopan and Tabita unpublished). In

this study, however, it was shown that the I165T enzyme was able to support anaerobic

MTA-dependent growth in the R. rubrum cbbM-/RLP- strain. Residue I165 is in Van der

Waals contact with two other active site residues, K191 and D193, and magnesium ions.

The alteration of residue 165 probably changed its ability to interact with other residues

important for catalysis and in turn most likely altered the binding of the RuBP substrate.

87 Clearly, as shown by the in vivo complementation studies, the interaction of these

residues may not be critical for binding of the substrate used in MTA metabolism.

Further work on the exact reaction catalyzed by RubisCO in anaerobic MTA metabolism

will surely contribute to a future understanding of how different residues interact with the

distinct substrates used by RubisCO.

With respect to the reaction catalyzed by R. rubrum RLP, recombinant protein

was purified after expressing the RLP gene in E. coli. This RLP, like other RLPs (Hanson and Tabita 2001; Ashida et al. 2003), did not catalyze either the carboxylation or

oxygenation of RuBP. This was expected based on the sequence analysis of RLP, which

had low similarity with bona fide RubisCO and contains many nonconservative active

site amino acid substitutions [7 out of 19 known residues (Figure 1.3)]. Based on the in

vivo results, it was expected that like B. subtilis RLP, R. rubrum RLP would also catalyze

enolization of DK-MTP 1-P as a part of the MSP. Because of the unstable nature of DK-

MTP 1-P, a substrate analog, 2, 3 ketohexane 1-phosphate (DK-H 1-P) (Zhang et al.,

2004) was used to test the ability of R. rubrum RLP to catalyze the enolase reaction. DK-

H 1-P was originally described as the substrate analog of the enolase phosphatase enzyme of the MSP from Klebsiella oxytoca (Zhang et al., 2004). It was later shown that

RLP from Geobacillus kaustophilus could also catalyze the enolization of DK-H 1-P

(Imker et al., 2007). R. rubrum RLP failed to catalyze enolization of DK-H 1-P, yet control experiments indicated that B. subtilis RLP was fully active with this substrate. In a collaborative study with Dr. John A Gerlt’s group at the University of Illinois at Urbana-

Champaign, the actual substrate, DK-MTP 1-P was synthesized and used to determine whether it is the analog that is not recognized by the R. rubrum RLP or whether the RLP does not have the ability to perform the enolase reaction. NMR studies using

88 enzymatically synthesized DK-MTP 1-P showed that R. rubrum RLP actually utilized

MTRu 1-P as its substrate and catalyzed an unusual isomerization reaction (Figure

3.14). Exactly how this reaction ties in with an altered MSP and fits in with MTA metabolism in vivo remains to be determined. Genes encoding for some novel enzymes may be present in the genome which are involved in the conversion of the RLP product

1-methylthio-xylulose/ribulose 1-phosphate back to methionine.

CHAPTER 4

PHOTOAUTOTROPHIC GROWTH OF A Rhodospirillum rubrum RUBISCO MUTANT

INTRODUCTION

Rhodospirillum rubrum and Rhodobacter sphaeroides fix CO2 by means of the

CBB cycle, in which RubisCO is the key CO2 fixing enzyme. A RubisCO disruption strain of R. rubrum and a double RubisCO disruption strain of R. sphaeroides are not capable of growing under photoautotrophic (PA) condition using hydrogen as the electron source.

However, the R. rubrum RubisCO disruption strain can grow photoheterotrophically using organic compounds such as malate as a carbon source whereas the R. sphaeroides RubisCO disruption strain cannot (Falcone and Tabita, 1993; Wang et al.,

1993a). RubisCO mutant strains from both organisms were shown to grow under photoautotrophic condition when less reduced electron donors, thiosulfate or sulfide, were used as electron donors instead of hydrogen. Growth in the absence of RubisCO suggested the presence of another, independent CO2 fixing pathway (Wang et al.,

1993b).

90 Genome analysis of R. rubrum showed the absence of key enzymes of known

CO2 fixing pathways (Table 4.1). This suggested that it is possible that none of these

CO2 fixing pathways are functional in R. rubrum. However, the possibility of

nonhomologous proteins catalyzing key reactions of the alternate CO2 fixation pathways

cannot be ruled out at this time.

R. rubrum Pathway Enzyme Organism homolog 3-OH propionate/malyl- Chloroflexus Rru_A3667; Malonyl-CoA CoA cycle, 3-OH aurantiacus 29% identical reductase propionate/4-OH Metallosphaera sedula None butyrate cycle 3-OH propionate/4-OH 4-hydroxybutyrate Metallospaera sedula None butyrate cycle dehydratase Hydrogenobacter None ATP citrate lyase thermophilus Reductive TCA cycle

C. tepidum None Reductive acetyl-CoA Acetyl-CoA Moorella thermoacetica None pathway synthase

Table 4.1: Summary of a genome search showing the absence of homologs of the key enzymes of known CO2 fixing pathways in R. rubrum.

A novel pathway of acetate assimilation, the ethylmalonyl-CoA pathway, was recently described in R. sphaeroides (Figure 4.1) (Alber et al., 2006 and Erb et al.,

2007). Genome analysis of R. rubrum shows the presence of a homolog of the genes encoding key enzymes of the pathway, namely, ethylmalonyl-CoA mutase (Rru_A3062,

Erb et al., manuscript submitted), methylsuccinyl-CoA dehydrogenase (Rru_A30,

91 Zarzycki et al., 2008), mesaconyl-CoA hydratase (Rru_A3064), malyl-CoA/β-

methylmalyl-CoA lyase (Rru_A0217, Meister et al., 2005) and the CO2 fixing enzyme crotonyl-CoA carboxylase reductase (ORF, Rru_A3063). The possibility of the presence of a functional ethylmalonyl-CoA pathway in R. rubrum was thus investigated. Crotonyl-

CoA carboxylase-reductase (Ccr), the key CO2 fixing enzyme of the pathway catalyzes

the following reaction:

- + crotonyl-CoA + NADPH + CO2 ethylmalonyl-CoA + NADP

92 O O acetyl-CoA acetyl-CoA

H 3 C S-CoA H 3 C S-CoA O O acetoacetyl-CoA

H3C S-CoA

OH O (R)-3-hydroxybutyryl-CoA H3C S-CoA

O crotonyl-CoA

H3C S-CoA CO2 O ethylmalonyl-CoA

H3C S-CoA COOH O methylsuccinyl-CoA HOOC S-CoA

CH3 O mesaconyl-CoA

HOOC S-CoA

CH3 β-methylmalyl-CoA OH O O acetyl-CoA HOOC S-CoA HCO O O H3C S-CoA CH3 Succi- H3C malate nate S-CoA HOOC H propionyl-CoA glyoxylate

Figure 4.1: The ethylmalonyl-CoA pathway as studied in isocitrate lyase-negative organisms such as the nonsulfur purple bacterium R. sphaeroides. Crotonyl-CoA carboxylase/reductase was identified as the key CO2 fixing enzyme (Modified from Erb et al., 2007).

93 MATERIALS AND METHODS

Bacterial strains and growth conditions. The R. rubrum strains used in the

current study are Str-2 (wild type, a spontaneous streptomycin (Sm) resistant derivative

of strain S1 (ATCC 11170) and strain I-19 (cbbM-) (RubisCO disruption strain) (Falcone and Tabita, 1993). Thiosulfate- and sulfide-dependent growth was achieved by using autotrophic PF-7 medium supplemented with biotin (15 µg/liter), as previously described in this study [Chapter 2; (Mukhopadhyay et al., 1999; Wahlund and Madigan 1993, 1995 and Wahlund et al., 1991)]. Acetate and succinate were used for photoheterotrophic

growth.

Preparation of cell extracts for enzyme assays. Cells (20 ml culture) were

harvested at an Optical density at 660 nm of ~ 0.4 by centrifugation at 7500 G for 15

minute, washed with TEM buffer (20 mM Tris-HCl, 10 mM MgCl2, 1 mM EDTA, 50 mM

NaHCO3, 5% glycerol, pH 8) and suspended in 250 µl of TEM. Glass beads were added

to the cell suspension and the cells then lysed using a cell mill MM200 (Retsch, PA,

USA). Lysates were then centrifuged to separate the glass beads. The Bradford method

was used to determine protein concentrations. BSA (bovine serum albumin) was used to

standardize the assay (Bradford, 1976).

Synthesis of crotonyl-CoA. Crotonyl-CoA was provided by Dr. B. E. Alber. It

was synthesized from its anhydrides: 0.336 gm NaHCO3 was dissolved in 40 ml anaerobic H2O and then 80 mg coenzyme A was added [performed inside an anaerobic

hood (Coy labs, Grass Lake, Michigan)]; the bottle was then sealed with rubber stoppers

with an aluminum seal crimped over the stopper (Bellco Glass Inc. Vineland, NJ, USA).

Outside the anaerobic chamber, 22 µl of crotonic anhydride was added to it using a

94 syringe through the rubber septa. The reaction was incubated while stirring at room

temperature for 1 h. After 1 h the pH of reaction was adjusted to 3.0 with H2SO4 and

crotonyl-CoA was extracted with diethylether (two times). The aqueous phase,

containing crotonyl-CoA, was lyophilized to obtain crotonyl-CoA (Simon and Shemin,

1953).

Radiometric assay for crotonyl-CoA carboxylase/reductase (Ccr). Ccr

14 - activity was determined by following the incorporation of [ C]-HCO3 into acid stable ethylmalonyl-CoA. The reaction mixture (0.5 ml) contained 80 mM Tris-HCl buffer (pH

14 8.0), 3 mM crotonyl-CoA, 3 mM NADPH, 5 µC/ml NaH CO3, and cell extract (0.2 - 0.4

µg ml-1). The reaction was performed at 30 C and was initiated by adding NADPH to the

assay mixture. The reaction was then stopped at different time points by transferring 100

µl of the reaction mixture into a tube containing 50 µl propionic acid. The samples were

14 14 centrifuged for 30 min to remove nonincorporated CO2 and the amount of fixed C was determined by liquid scintillation counting after mixing 100 µl of the reaction mixture with

3 ml scintillation cocktail and counting. Background radioactivity was determined after performing reactions in the absence of NADPH and crotonyl-CoA.

95 RESULTS

Growth of R. rubrum on thiosulfate medium. R. rubrum strains were grown on

PF-7 medium (described elsewhere for the growth of C. tepidum). Thiosulfate (with a small amount of sulfide added to keep the medium reduced) was found to support autotrophic growth. Using the PF-7 medium, the maximum turbidity achieved

(OD660~0.7) was considerably less than that obtained with photoautotrophic growth in

Ormerod’s medium using hydrogen as the electron donor (OD660~2.0). It was also clearly

shown that strain cbbM-, the RubisCO mutant, grows under photoautotrophic conditions using thiosulfate and possibly sulfide as electron donors.

The R. rubrum wild type and RubisCO disruption strains were also grown in PF-7 medium under photoheterotrophic growth conditions, using acetate and succinate as the carbon sources. Both the wild type and cbbM- strain exhibited lags in growth when cultured in the PF-7 medium, the lag was observed in cultures grown using both photoheterotrophically (Ormerod’s medium) grown as well as photoautotrophically (PF-7 medium) grown cells. The cbbM- strain showed more of a lag in growth (200-250 h)

compared to the wild type strain (100 - 150 h) (Figure 4.2).

96 A

1.5

m 1

OD 660 n

0.5

0 0 100 200 300 400 Time (hours)

1.5

OD 660 n OD 0.5

0 100 200 300 400

Time (hours)

Figure 4. 2. Growth of R. rubrum strains on PF-7 medium using thiosulfate/sulfide as electron donors. (A), wild type strain; (B), cbbM-. Carbon sources, bicarbonate (▲), acetate (∆) and succinate (X).

97 The absence of the RubisCO protein was confirmed in the cbbM- strain grown on

PF-7 medium under photoautotrophic growth conditions after performing Western

immunoblot analyses using antisera to the R. rubrum RubisCO (Figure 4. 3).

1 2 3

50 K.D.

Figure 4.3: Western immunoblot using R. rubrum RubisCO antibody showing the absence of RubisCO protein in extracts from the cbbM- strain grown under photoautotrophic conditions. Lanes: (1), purified R. rubrum RubisCO; (2), R. rubrum wild type; and (3), R. rubrum cbbM- strain. .

98 Crotonyl CoA carboxylase/reductase activity in R. rubrum cell extracts. The activity for the enzyme crotonyl-CoA carboxylase/reductase was determined by a radiometric assay. Cells grown under photoautotrophic conditions using thiosulfate/sulfide as the electron donors clearly showed Ccr activity (Table 4.1).

Ccr Specific activity (nmoles. mg-1.min-1) Carbon source

WT cbbM-

Autotrophic 169 77

Acetate 89 10

Succinate 54 9

Table 4.2: Ccr activity in the R. rubrum strains grown on medium containing thiosulfate and sulfide. WT; wild type and cbbM- (RubisCO disruption strain). Result of a single experiment.

99 DISCUSSION

In a previous study it had been demonstrated that a RubisCO disruption strain of

R. rubrum, lacking the CBB cycle could grow under photoautotrophic conditions using

CO2 as the carbon source, when thiosulfate and sulfide were used as the electron

donors (Wang et al., 1993b). However, this strain could not grow using CO2 as sole

carbon source when hydrogen was used as the electron donor. Based on these results it

was hypothesized that two independent pathways of CO2 fixation might be present in R.

rubrum. The potential presence of unknown alternative pathway(s) of CO2 fixation in

nonsulfur purple bacteria (R. rubrum and R. sphaeroides) had been suggested earlier.

(Wang et al., 1993b). But why these organisms might use an alternative CO2 fixation pathway when grown on reduced sulfur sources, in the absence of RubisCO, is not known at this time. Perhaps, since thiosulfate and sulfide have higher redox potentials compared to hydrogen, this somehow favors the use of the alternative pathway (s). Both the wild type and RubisCO disruption strain were able to grow on PF-7 medium containing thiosulfate as the main electron donor (in the presence of a small amount of sulfide) to similar final cell densities (ODe660~ 0.6 – 0.7). However, the RubisCO mutant

strain exhibited a longer lag under these growth conditions. In all cases, growth on PF-7

medium supplemented with acetate or succinate was significantly higher compared to

growth in the autotrophic PF-7 medium. The absence of the RubisCO protein was

confirmed after performing Western immunoblot analyses of extracts from the R. rubrum

RubisCO disruption strain, excluding the possibility of any wild type contamination in

these experiments. The use of PF-7 media under rigorously established anaerobic

100 growth conditions greatly improved photoautotrophic growth of both the mutant and wild-

type strains compared to the previous study (Wang et al., 1993b).

After establishing good growth of the RubisCO disruption strain under

photoautotrophic conditions, the presence of the recently identified ethylmalonyl-CoA

pathway (Alber et al., 2006, Erb et al., 2007) for acetate assimilation was investigated.

The activity of one of the key CO2 enzyme of this pathway, crotonyl-CoA carboxylase/reductase (Ccr), which catalyzes the conversion of crotonyl-CoA to ethylmalonyl-CoA, was assayed in cell extracts from the R. rubrum strains. Extracts from both the wild type and the cbbM- strain showed Ccr activity. It was surprising to find Ccr

activity under these growth conditions, as it is not obvious why the cells may require a

functional ethylmalonyl-CoA pathway during photoautotrophic growth. It is possible that

the ethylmalonyl-CoA pathway is constitutively expressed in R. rubrum. Alternately, the

ethylmalonyl-CoA pathway maybe upregulated in the presence of thiosulfate/sulfide.

Studies involving cells grown under photoautotrophic conditions with H2 as electron

donor will be required to answer this question. Regardlessly, for R. rubrum to be able to grow autotrophically utilizing one of these acetate assimilation pathways, the means by which this organism synthesizes acetate under autotrophic conditions in the absence of the CBB cycle also needs to be investigated. Despite the fact that we have unequivocally shown that R. rubrum is capable of substantial photoautotrophic growth in the absence of the CBB pathway, much more work is required to show that one of these pathways might be involved in photoautotrophic growth of this organism.

101

CHAPTER 5

RECAPITULATION AND FUTURE DIRECTIONS

SUMMARY OF WORK PERFORMED THUS FAR

Form I, II and III RubisCO are bona fide RubisCOs, capable of catalyzing RuBP- dependent carboxylation and oxygenation with different substrate specificities under different physiological contexts. The discovery of the RubisCO homolog, the RubisCO- like protein (RLP) (form IV RubisCO), in diverse organisms such as Chlorobium tepidum

and Bacillus subtilis that lack a functional CBB cycle , not only raised the question as to

why a RubisCO homolog is even present in these organisms but also whether such

RubisCO homologs have any specific functional relevance. Further studies involving in

vitro characterization and mutations in the RLP genes in these organisms showed that

although they lack the ability to perform a bona fide RubisCO reaction; i.e. carboxylation

and/or oxygenation of RuBP, they still play important roles in their respective organisms.

Based on phylogenetic analyses, the RLPs represent a special class of RubisCO; they

are similar to bona fide RubisCOs in overall structure but have enough differences at the primary sequence and structural level to make them incapable of catalyzing a typical

RubisCO reaction. RLPs have further been divided into six different subgroups; IV-

Photo, IV-YkrW, IV-DeepYkrW, IV-NonPhoto, IV-AMC and IV-GOS. Close analysis of the RLP sequences from the different subgroups or clades suggest that many of these

102 proteins may not be performing the same function. In fact, studies with C. tepidum RLP,

a member of the IV-Photo group, and a few members of the IV-YkrW group, indicate that

these RLPs are performing functions distinct from bona fide RubisCO as well as each other. The exact function of C. tepidum RLP is not known, but there is a strong indication of its involvement in thiosulfate oxidation, whereas RLP members of the IV-YkrW group have been shown to act as enolases in the methionone salvage pathway (MSP).

As mentioned, previous studies with an RLP disruption strain of C. tepidum

indicated the role of this RLP in thiosulfate oxidation; the mutant strain also showed

defects with respect to photoautotrophic growth, pigmentation, elemental sulfur oxidation

and increased levels of two oxidative stress response proteins. The high levels of

oxidative stress response proteins in the RLP disruption strain of C. tepidum enables the organism to tolerate and exhibit a high level of hydrogen peroxide resistance compared to the wild type. The first part of the current study was an extension of previous work done with the RLP mutant in C. tepidum. Investigations were initiated in order to discern the nature of the signal elicited by disruption of RLP, which leads to overexpression of genes encoding two oxidative stress response proteins, Tsa (thiol specific antioxidant protein) and SOD (superoxide dismutase). Experiments were conducted to identify likely transcriptional regulator(s) of the tsa and sod genes. Fur (Ferric uptake regulator) family members are known to be involved in the regulation of these genes in other organisms.

Two members of this family, Fur and PerR, are encoded by the C. tepidum genome,

Their involvement in the regulation of tsa and sod gene transcription was investigated by constructing fur and perR disruption strains both independently and in conjunction with the RLP disruption. Analysis of these mutant strains indicated that the Fur and PerR proteins were involved in regulating some other genes in the organism, but both the tsa

103 and sod genes were not apparently under their control, and there was no direct involvement of RLP with the Fur and PerR regulators.

Functional complementation of the RLP disruption strain was attempted by expressing the RLP gene on a new complementation vector created in this lab by Dr.

Thomas Hanson. The vector containing the RLP gene was successfully transferred into

C. tepidum by conjugation using E. coli S17λpir as the donor strain. Immunoblot analysis

using the anti RLP antibody showed successful expression of the RLP gene. However,

the RLP disruption strain Ω::RLP, with the RLP gene expressed on the plasmid, still

showed the pigmentation defect and this strain still overexpressed the genes encoding

for the Tsa and SOD proteins. These results suggested that the RLP disruption caused

complex changes somewhere else in the genome/proteome, which were not reversed by

the expression of the RLP gene provided on the plasmid.

It had been shown earlier that the B. subtilis RLP acts as an enolase in the MSP.

The observation that form II RubisCO from R. rubrum can functionally complement for

the loss of RLP gave rise to the hypothesis that different members of the RubisCO family

might functionally complement each other. Another question raised by these results was

whether the R. rubrum RubisCO catalyzed an enolase reaction in its native environment

and for what purpose. Besides RubisCO’s ability to catalyze the enolase reaction of the

MSP, another interesting feature about R. rubrum is the fact that this organism contains both bona fide RubisCO and RLP. This makes R. rubrum an ideal system to study RLP

in its native organism and test for the ability of bona fide RubisCO and/or RLP to

functionally complement each other and catalyze distinct functions, while providing a

clear physiological context for such capabilities.

104 Prior to this study, the significance and role of a functional MSP had not been demonstrated in the nonsulfur purple bacteria. Growth studies using MTA as the sole sulfur source showed that there was a direct correlation between MTA-dependent growth and the expression of the RLP gene in this organism. Moreover, unlike

Rhodobacter spharoides and Rhodobacter capsulatus, the genome of

Rhodopseudomonas palustris was also shown to encode for functional RLP genes, and like R. rubrum, R. palustris also showed the ability to metabolize MTA. Both R. rubrum and R. palustris also proved valuable for other reasons as there have not been reports of a functional MSP under anaerobic growth conditions. The ability of these two nonsulfur purple bacteria to grow both under aerobic and anaerobic conditions gave rise to the possibility of investigating the presence and role of the MSP under anaerobic growth conditions. Experiments were thus performed to investigate the ability of these organisms, especially R. rubrum, to metabolize MTA under anaerobic conditions. It was found that these organisms, indeed, have the ability to metabolize MTA under anaerobic conditions.

To further study the role of the MSP in R. rubrum and R. palustris in MTA metabolism, we prepared RLP disruption strains. In R. rubrum, as expected, the RLP disruption strain lacked the ability to metabolize MTA under aerobic growth conditions, indicating the requirement of RLP to enable MTA metabolism. Surprisingly, the RLP disruption strain was able to use MTA as sole sulfur source when grown under anaerobic growth conditions. This suggested that either there is another protein performing the function normally provided by RLP under this growth condition or there is an alternative pathway for MTA utilization which does not require RLP. As mentioned earlier, it has been shown that RubisCO from R. rubrum is capable of catalyzing the in

105 vitro and in vivo enolization of DK-MTP 1-P in B. subtilis. This raised the possibility of

RubisCO’s involvement in MTA-dependent growth under anaerobic growth conditions, a

rather novel role for this enzyme as it is also needed for CO2 reduction under these growth conditions. Confirming the involvement of RubisCO in MTA utilization under anaerobic growth conditions, a RubisCO disruption strain of R. rubrum was unable to

grow using MTA as sole sulfur source., Moreover, an RLP and RubisCO double

disruption strain was constructed that lacked the ability to metabolize MTA both under

aerobic and anaerobic conditions, making it an ideal strain to perform complementation

studies to further probe the role of diverse RLP and RubisCO molecules.

As discussed above, the R. rubrum RubisCO disruption strain lacked the ability

to use MTA as the sole sulfur source under anaerobic conditions, however the double

RubisCO disruption strain of R. palustris was able to grow using MTA under these

conditions. Unlike R. rubrum, this suggested that one or both of the RLP molecules

might be able to perform this function under anaerobic conditions in this organism.

Growth studies using some of the RubisCO and RLP gene mutant strains of R. palustris

did not yield any conclusive information about the ability of the individual proteins to

participate in MTA metabolism.

Complementation studies were performed using the RubisCO/RLP disruption

strain both under aerobic and anaerobic conditions. B. subtilis RLP was able to partially

complement the RubisCO/RLP disruption strain under aerobic conditions. However, form

II RubisCO from R. palustris (RpcbbM) was not able to complement the double knockout

strain under aerobic growth conditions, most likely due to either failure of the R. palustris

cbbM gene to be expressed or perhaps due to degradation of the RubisCO protein.

Form II RubisCO from R. palustris was able to support MTA-dependent growth of the

106 RubisCO/RLP mutant strain of R. rubrum under anaerobic conditions. In addition, the R.

palustris CbbM protein containing an I165T alteration was able to support MTA-

dependent growth while the same mutant RubisCO was not able to support

photoautotrophic growth of a RubisCO deletion strain of R. capsulatus. These results

suggest that residue 165 is a residue important for the RuBP carboxylation reaction but

not the reaction used in MTA metabolism.

In vitro studies of R. rubrum RLP indicated that it lacked ability to catalyze the

enolase reaction when DK-H 1-P, a homolog of DK-MTP 1-P, was used as the

substrate. Further studies in collaboration with Prof. John Gerlt’s group at the University

of Illinois showed that this RLP also lacked DK-MTP 1-P enolase activity. Interestingly, it

was found that instead of DK-MTP 1-P, the R. rubrum RLP utilized MTRu-1-P (another

intermediate in the MSP) as a substrate and converted it into a novel compound which is

not an intermediate of the conventional MSP.

Finally, it was shown that the R. rubrum RubisCO disruption strain could grow

under photoautotrophic conditions using thiosulfate/ sulfide as electron donors,

indicating the presence of a pathway other than the CBB cycle for CO2 fixation. The possibility of the presence of the ethylmalonyl-CoA pathway of acetate assimilation was explored by testing for crotonyl-CoA carboxylase/reductase (Ccr) activity, a key enzyme in the pathway. R. rubrum cell extracts possesed Ccr activity, which preliminary results seemed to indicate that its level was lower in the RubisCO disruption strain compared to the wild type.

107 SUGGESTED FUTURE EXPERIMENTS

Complementation studies in R. rubrum involving different RLP and

RubisCO genes. Complementation studies involving more genes from different forms of

RubisCO will provide more information on the ability of those proteins to participate in

MTA metabolism. These studies will also help understand important features of the

RubisCO active site.

Further in vitro assays of the novel reaction catalyzed by R. rubrum RLP.

The novel reaction catalyzed by this RLP can be assayed by NMR methods. This technique can be exploited to directly assay the capabilities of different RubisCOs and

RLPs to carry out this reaction. These studies can be extended to mutant RubisCO and

RLP proteins, which can help understand the role of different active site residues in catalysis.

Mutation in the gene linked with R. rubrum RLP. Most of the RLPs within the

IV-DeepYkrW lineage seem to be linked on the genome to a protein of the cupin2 super family (Figure 5.1). In fact, the dioxygenase involved in MSP is also a member of the same super family, but the proteins linked to the IV-DeepYkrW RLPs are not homologous to the dioxygenases. The cupin2 super family is a very diverse family of proteins with members identified as dioxygenases, phopsphomannose isomerases, oxalate decarboxylase, etc. (Khuri et al., 2001). It is possible that these proteins are part of a novel pathway of MTA metabolism. Gene disruption in the cupin2 family member protein in R. rubrum should provide more information regarding its involvement in MTA metabolism.

108

Figure 5.1: Local conservation near genes encoding RLPs of the IV-DeepYkrW lineage showing conservation of hypothetical proteins of the Cupin2 super family next to the RLPs. Members of the cupin2 super family are indicated by a star. Gene neighborhoods were visualized using tools at the Integrated Microbial Genomes website (http://img.jgi.doe.gov/cgi-bin/pub/main.cgi). RLP genes are indicated in red. Other open reading frames are colored and identified according to their annotation in the Integrated Microbial Genomes database.

109 Creation of different mutant strains of RubisCO(s) and RLP(s) in R. palustris. The presence of two bona fide RubisCOs and two RLPs makes the characterization of individual proteins difficult with respect to their ability to participate in

MTA metabolism. There are major differences in R. palustris and R. rubrum with respect to MTA metabolism. Most importantly, in R. palustris one or both of the RLPs was able to support MTA-dependent growth in the absence of both RubisCOs, whereas the R. rubrum RubisCO disruption strain was incapable of MTA metabolism under anaerobic growth conditions. These observations warrant the construction of further strains with additional mutations in the RubisCO and RLP genes so that this observation might indicate whether the two distinct RLPs perform different functions in R. palustris. A quadruple disruption strain of both the RubisCOs and RLPs will be helpful in performing complementation studies to help resolve this issue.

Mutation in the ccr gene in R. rubrum. A disruption in the ccr gene will be helpful with respect to answering the question as to whether it is involved in the alternative autotrophic pathway of CO2 fixation in this organism. The initial step will be to test the ability of the ccr- strain to grow under photoautotrophic conditions using thiosulfate and sulfide as electron donors.

110

REFERENCES

Alber, B. E., R. Spanheimer, C. Ebenau-Jehle, and G. Fuchs. 2006. Study of an alternate glyoxylate cycle for acetate assimilation by Rhodobacter sphaeroides. Mol. Microbiol. 61: 297-309.

Arsene, F., P. A. Kaminski and C. Elmerich. 1996. Modulation of NifA activity by PII in Azospirillum brasilense: evidence for a regulatory role of the NifA N-terminal domain. J. Bacteriol. 178(16): 4830-4838.

Ashida, H., Y. Saito, C. Kojima, K. Kobayashi, N. Ogasawara and A. Yokota. 2003. A functional link between RuBisCO-like protein of Bacillus and photosynthetic RuBisCO. Science. 302 (5643): 286-290.

Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith & K. Struhl. (1987) Current Protocols in Molecular Biology (Green & Wiley, New York).

Bassham, J. A., A. A. Benson, M. Calvin. 1950. The path of carbon in photosynthesis. J Biol Chem. 185(2):781-787.

Berg, I.A., L. V. Filatova and R. N. Ivanovsky. 2002. Inhibition of acetate and propionate assimilation by itaconate via propionyl-CoA carboxylase in isocitrate lyase- negative purple bacterium Rhodospirillum rubrum. FEMS Microbiol Lett. 216(1): 49-54.

Berg, I. A., D. Kockelkorn, W. Buckel and G. Fuchs. 2007. A 3-hydroxypropionate/4 hydroxybutyrate autotrophic carbon dioxide assimilation pathway in Archaea. Science. 318(5857): 1782-1786.

Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248- 254.

Bult, C. J., O. White, and G. J. Olsen et al. 1996. Complete genome sequence of the methanogenic archeon, Methanococcus jannaschii. Science 273:1058-1073.

Carré-Mlouka A, A. Méjean, P. Quillardet, H. Ashida, Y. Saito, A. Yokota, I. Callebaut, A. Sekowska, E. Dittmann, C. Bouchier and N. T. de Marsac. 2006. A new

111 rubisco-like protein coexists with a photosynthetic rubisco in the planktonic cyanobacteria Microcystis. J. Biol. Chem. 281(34): 24462-24471.

Chan L. K., T. S. Weber, R. M. Morgan-Kiss, T. E. Hanson. 2008. A genomic region required for phototrophic thiosulfate oxidation in the green sulfur bacterium Chlorobium tepidum (syn. Chlorobaculum tepidum). Microbiology. 2008 Mar;154(Pt 3):818-829.

Chène, P., A. G. Day and A. R. Fersht . 1997. Role of isoleucine-164 at the active site of rubisco from Rhodospirillum rubrum. Biochem Biophys Res Commun. 232(2): 482- 486.

Chung, S., G. Shen, J. Ormerod & D. A. Bryant. 1998 FEMS Microbiol. Lett.164, 353– 361.

Cleland, W. W. T. J. Andrews, S. Gutteridge, F. C. Hartman, and G. H. Lorimer. 1998. Mechanism of RuBisCO: The Carbamate as General Base, Chem. Rev. 98: 549-562.

Cook, L. S., H. Im and F. R. Tabita. 1988. Oxygen-dependent inactivation of ribulose 1,5-bisphosphate carboxylase/oxygenase in crude extracts of Rhodospirillum rubrum and establishment of a model inactivation system with purified enzyme. J. Bacteriol. 170(12): 5473-5478.

Cook, L. S. and F. R. Tabita. 1988. Oxygen regulation of ribulose 1, 5-bisphosphate carboxylase activity in Rhodospirillum rubrum. J Bacteriol. 170(12): 5468-5472.

Derelle E, C. Ferraz, S. Rombauts, P. Rouzé, A. Z. Worden, S. Robbens, F. Partensky, S. Degroeve, S. Echeynié, R. Cooke, Y. Saeys, J. Wuyts, K. Jabbari, C. Bowler, O. Panaud, B. Piégu, S. G. Ball, J. P. Ral, F. Y. Bouget, G. Piganeau, B. De Baets, A. Picard, M. Delseny, J. Demaille, Y. Van de Peer and H Moreau. 2006. Genome analysis of the smallest free-living eukaryote Ostreococcus tauri unveils many unique features. Proc Natl Acad Sci U S A. 103 (31): 11647-11652.

Dower, W. J., J. F. Miller and C. W. Ragsdale. 1988. High efficiency transformation of E. coli by high voltage electroporation. Nucleic Acids Res. 16(13): 6127-6145.

Erb, T. J., I. A. Berg, V. Brecht, M. Müller, G. Fuchs and B. E. Alber. 2007. Synthesis of C5-dicarboxylic acids from C2-units involving crotonyl-CoA carboxylase/reductase: the ethylmalonyl-CoA pathway. Proc. Natl. Acad. Sci. U. S. A. 104(25):10631-10636.

Evans, M. C., B. B. Buchanan and D. I. Arnon. 1966. A new ferredoxin-dependent carbon reduction cycle in a photosynthetic bacterium. Proc. Natl Acad Sci U S A. 55 (4): 928-934.

Falcone, D. L. and F. R. Tabita. 1993. Complementation analysis and regulation of CO2 fixation gene expression in a ribulose 1,5 bisphosphate carboxylase-oxygenase deletion strain of Rhodospirillum rubrum. J. Bacteriol. 175: 5066-5077.

112 Finn, M. W. and F. R. Tabita. 2004. Modified pathway to synthesize ribulose 1,5- bisphosphate in methanogenic archaea. J. Bacteriol. 186(19): 6360-6366.

Frangeul, L., P. Quillardet, A. M. Castets, J. F. Humbert, H. C. Matthijs, D. Cortez, A. Tolonen, C. C. Zhang, S. Gribaldo, J. C. Kehr, Y. Zilliges, N. Ziemert, S. Becker, E. Talla, A. Latifi, A. Billault, A. Lepelletier, E. Dittmann, C. Bouchier, N. T. de Marsac. 2008. Highly plastic genome of Microcystis aeruginosa PCC 7806, a ubiquitous toxic freshwater cyanobacterium. BMC Genomics.9:274.

Galagan, J. E., C. Nusbaum, A. Roy, and M. G. Endrizzi et al. 2002. The genome of M. acetivorans reveals extensive metabolic and physiological diversity. Genome Res. 12: 532-542.

Hanson, T. E. and F. R. Tabita. 2001. A ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO)-like protein from Chlorobium tepidum that is involved with sulfur metabolism and the response to oxidative stress. Proc Natl Acad Sci U S A. 98(8): 4397-4402.

Hanson, T. E. and F. R. Tabita. 2003. Insights into the stress response and sulfur metabolism revealed by proteome analysis of a Chlorobium tepidum mutant lacking the Rubisco-like protein. Photosyn. Res. 78: 231-248.

Herbig, A. F. and J. D. Helmann. 2001. Roles of metal ions and hydrogen peroxide in modulating the interaction of the Bacillus subtilis PerR peroxide regulon repressor with operator DNA. Mol. Microbiol. 41(4): 849-859.

Herbig, A. F., and J. D. Helmann. 2002. Metal ion uptake and oxidative stress, p. 405- 414. In A. L. Sonenshein, J. A. Hoch, and R. Losick (ed.), Bacillus subtilis and its closest relatives: from genes to cells. ASM Press, Washington, D.C.

Horsburgh, M. J., M. O. Clements, H. Crossley, E. Ingham, S. J. Foster. 2001. PerR controls oxidative stress resistance and iron storage proteins and is required for virulence in Staphylococcus aureus. Infect. Immun. 69(6): 3744-3754.

Imker, H.J., A. A. Fedorov, E. V. Fedorov, S. C. Almo and J. A. Gerlt. 2007. Mechanistic diversity in the RuBisCO superfamily: the "enolase" in the methionine salvage pathway in Geobacillus kaustophilus. Biochemistry. 46(13): 4077-4089.

Keen, N. T., S. Tamaki, D. Kobayashi and D. Trollinger. 1988. Improved broad-host- range plasmids for DNA cloning in Gram negative bacteria. Gene 70: 191-197.

Khuri, S., F. T. Bakker and J. M. Dunwell. 2001. Phylogeny, function, and evolution of the cupins, a structurally conserved, functionally diverse superfamily of proteins. Mol. Biol. Evol. 18(4): 593-605.

Klenk, H.P., R. A. Clayton, H. P. Klenk, R. A. Clayton, J. F. Tomb, O. White, K. E. Nelson, K. A. Ketchum, R. J. Dodson, M. Gwinn, E. K. Hickey, J. D. Peterson, D. L. Richardson, A. R. Kerlavage, D. E. Graham, N. C. Kyrpides, R. D. Fleischmann, J. 113 Quackenbush, N. H. Lee, G. G. Sutton, S. Gill, E. F. Kirkness, B. A. Dougherty, K. McKenney, M. D. Adams, B. Loftus, S. Peterson, C. I. Reich, L. K. McNeil, J. H. Badger, A. Glodek, L. Zhou, R. Overbeek, J. D. Gocayne, J. F. Weidman, L. McDonald, T. Utterback, M. D. Cotton, T. Spriggs, P. Artiach, B. P. Kaine, S. M Sykes, P. W. Sadow, K. P. D'Andrea, C. Bowman, C. Fujii, S. A. Garland, T. M. Mason, G. J. Olsen, C. M. Fraser, H. O. Smith, C. R. Woese and J. C. Venter. 1997. The complete sequence of the hyperthermophilic sulfate-reducing archaeon Archaeoglobus fulgidus. Nature. 390: 364-370.

Larimer, F. W., P. Chain, L. Hauser, J. Lamerdin, S. Malfatti, L. Do, M. L. Land, D. A. Pelletier, J. T. Beatty, A. S. Lang, F. R. Tabita, J. L. Gibson, T. E. Hanson, C. Bobst, J. L. Torres , C. Peres, F. H. Harrison, J. Gibson J and C. S. Harwood. 2004. Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palustris. Nat. Biotechnol. 21: 55-61.

Laemmli, U. K. 1970. Cleaving of structural proteins during the assembly during the assembly of the head of bacteriophage T4. Nature 227: 680-685.

Li, H., M. R. Sawaya, F. R. Tabita and D. Eisenberg. 2005. Crystal structure of a RuBisCO-like protein from the green sulfur bacterium Chlorobium tepidum. Structure. 13(5): 779-789.

Markwell, M. A., S. M. Haas, L. L. Bieber and N. E. Tolbert. 1978. A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Anal. Biochem. 87: 206–210.

Mukhopadhyay, B., E. F. Johnson, F. & M. J. Ascano. 1999 Appl. Environ. Microbiol. 65, 301–306.

Meister, M., S. Saum, B. E. Alber and G. Fuchs. 2005. L-malyl-coenzyme A/beta- methylmalyl-coenzyme A lyase is involved in acetate assimilation of the isocitrate lyase- negative bacterium Rhodobacter capsulatus. J. Bacteriol. 187(4):1415-1425.

Murphy, B. A., F. J. Grundy and T. M. Henkin. 2002. Prediction of gene function in methylthioadenosine recycling from regulatory signals. J. Bacteriol. 184(8): 2314-2318.

Ormerod, J. G., K. S. Ormerod and H. Gest. 1961. Light-dependent utilization of organic compounds and photoproduction of molecular hydrogen by the photosynthetic bacteria; relationships with nitrogen metabolism. Arch. Biochem. Biophys. 94: 449-463.

Paoli, G. C., P. Vichivanives, F. R. Tabita. 1998. Physiological control and regulation of the Rhodobacter capsulatus cbb operons. J. Bacteriol. 180(16):4258-69

Papworth, C., J. C. Bauer, J. Braman and D. A. Wright. 1996. Site-directed mutagenesis in one day with greater than 80% efficiency. Strategies. 9: 3-4.

114 Pearce, F.G. and T. J. Andrews . 2003. The relationship between side reactions and slow inhibition of ribulose-bisphosphate carboxylase revealed by a loop 6 mutant of the tobacco enzyme. J. Biol. Chem. 278(35): 32526-32536.

Penfold. R. J. and J. M. Pemberton. 1992. An improved suicide vector for construction of chromosomal insertion mutations in bacteria. Gene. 118(1):145-6.

Romagnoli, S. and F. R. Tabita. 2007. Phosphotransfer reactions of the CbbRRS three-protein two- component system from Rhodopseudomonas palustris CGA010 appear to be controlled by an internal molecular switch on the sensor kinase. J. Bacteriol. 189(2):325-335.

Sato, T., H. Atomi and T. Imanaka. 2007. Archaeal type III RuBisCOs function in a pathway for AMP metabolism. Science. 315(5814): 1003-1006.

Seiflein, T. A. and J. G. Lawrence. 2001. Methionine-to-cysteine recycling in Klebsiella aerogenes. J. Bacteriol. 183(1): 336-46.

Sekowska A., V. Dénervaud, H. Ashida, K. Michoud, D. Haas, A. Yokota and Danchin A. 2004. Bacterial variations on the methionine salvage pathway. BMC. Microbiol. 4; 4-9.

Simon, E. J. and D. Shemin. 1953 The Preparation of S-Succinyl Coenzyme A. J. of the American Chemical society. 75 (10): 2520 – 2520.

Simon, R., U. Priefer, and A. Puhler. 1983. A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in gram negative bacteria. Bio/Technology 1:784-791.

Smith, S. and F. R. Tabita. 2003. Positive and negative selection of mutant forms of prokaryotic (cyanobacterial) ribulose-1, 5-bisphosphate carboxylase/oxygenase. J. Mol. Biol. 331:557-569.

Tabita, F. R. 1999. Microbial ribulose 1, 5-bisphosphate carboxylase/oxygenase: A different perspective. Photosyn. Res. 60:1-28.

Tabita, F. R., T. E. Hanson, H. Li, S. Satagopan, J. Singh, and S. Chan. 2007. Function, structure, and evolution of the RubisCO-like proteins and their RubisCO homologs. Microbiol Mol Biol Rev. 71 (4):576-599.

Taylor, L. A. and R. E. Rose. 1988. A correction in the nucleotide sequence of the Tn903 kanamycin resistance determinant in pUC4K. Nucleic Acids Res. 16(1):358.

Tichi, M. A. and F. R. Tabita. 2002. Metabolic signals that lead to control of CBB gene expression in Rhodobacter capsulatus. J. Bacteriol. 184(7):1905-1915.

115 Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA. 76: 4350-4354.

Van Vliet, A. H., M. L. Baillon, C. W. Penn, and J. M. Ketley. 1999. Campylobacter jejuni contains two fur homologs: characterization of iron-responsive regulation of peroxide stress defense genes by the PerR repressor. J. Bacteriol. 181(20):6371-6.

Wahlund, T. M., C. R. Woese, R. W. Castenholz and M. T. Madigan. 1991. A thermophilic green sulfur bacterium from New Zealand hot springs, Chlorobium tepidum sp-nov. Arch. Microbiol. 156: 81–90.

Wang, X., D. L. Falcone and F. R. Tabita. 1993a. Reductive pentose phosphate- independent CO2 fixation in Rhodobacter sphaeroides and evidence that ribulose bisphosphate carboxylase/oxygenase activity serves to maintain the redox balance of the cell. J. Bacteriol. 175(11):3372-3379.

Wang, X., H. V. Modak and F. R. Tabita. 1993b. Photolithoautotrophic growth and control of CO2 fixation in Rhodobacter sphaeroides and Rhodospirillum rubrum in the absence of ribulose bisphosphate carboxylase-oxygenase. J. Bacteriol. 175(21):7109- 7114.

Weiner, M. P., C. Anderson C, B. Jerpseth, S. Wells, B. Johnson-Browne, P. Vaillancourt. 1994. Studier pET system vectors and hosts. Strateg. Mol. Biol. 7: 41–43.

Zhang, Y., M. H. Heinsen, M. Kostic, G. M. Pagani, T. V. Riera, I. Perovic, L. Hedstrom, B. B. Snider and T. C. Pochapsky. 2004. Analogs of 1-phosphonooxy-2, 2- ihydroxy-3-oxo-5- (methylthio)pentane, an acyclic intermediate in the methionine salvage pathway: A new preparation and characterization of activity with E1 enolase/phosphatase from Klebsiella oxytoca. Bioorg. Med. Chem. 12: 3847-3855.

116